Innovations in Effective Compassion

Compendium of Research Papers presented at the

Faith-Based and Community Initiatives Conference on Research, Outcomes, and Evaluation

Prepared for: U.S. Department of Health and Human Services

Office of the Assistant Secretary for Planning and Evaluation The Center for Faith-Based and Community Initiatives

Edited by: Pamela Joshi, PhD

Stephanie Hawkins, PhD Jeffrey Novey, MPH

RTI International

June 2008

This report is available on the Internet at: http://aspe.hhs.gov/fbci/comp08/

The views presented in these papers are those of the authors and do not necessarily represent the official position of the U.S. Department of Health and Human Services.

ACKNOWLEDGMENTS

The editors would like to thank the individuals at the universities, professional associations, think tanks, and federal agencies who distributed the calls for paper abstracts as the first step in identifying research papers to commission for this compendium. We also wish to acknowledge the helpful guidance of the Expert Panel members who reviewed numerous abstracts and recommended papers to invite.

Many individuals reviewed the papers for this research volume and we would like to thank our colleagues at RTI International who reviewed papers in their areas of expertise. In addition, staff members at the U.S. Department of Health and Human Services’ Office of the Assistant Secretary for Planning and Evaluation and the Center for Faith-Based and Community Initiatives, and at the White House Office of Faith-Based and Community Initiatives, reviewed the papers and provided helpful feedback.

Finally, we would like to acknowledge the contribution of the RTI editorial and production staff, including Justin Faerber and Valerie Garner.

The Compendium of Research Papers for the Faith-Based and Community Initiatives Conference was produced by RTI International* under Department of Health and Human Services contract No. 233020090.

Additional information about this project can be obtained from Alana Landey, the Federal Project Officer, who can be reached at alana.landey@hhs.gov or (202) 401-6636.

_______________________________________ *RTI International is a trade name of Research Triangle Institute.

CONTENTS

Introduction......................................................................................................................................1

Not By Government Nor Faith Alone: Rethinking Prisoner Reentry............................................17 Byron Johnson

Improving Organizational Capacity of Faith-Based and Community-Based Domestic Violence Service Providers ...........................................................................................39

Mitchell Brown

Partnerships That Work: Preliminary Findings from the TANF Faith-Based Community Organization Initiative ...............................................................................................61

Courtney J. Barthle, Rebecca M. Makar and Jeanette M. Hercik

Accessibility and Stability of Nonprofit Service Providers: Faith-Based and Community-Based Organizations in Urban and Rural America ...................................................79

Scott W. Allard

The Emerging Role of Faith-Based Organizations in the Low-Income Housing Market ..........................................................................................................................................103

Richard Hula, Cynthia Jackson-Elmoore and Laura Reese

Opening the Black Box of Faith-Based Services: Measuring Program Components and Treatment Dose................................................................................................147

Fred J. De Jong and Claudia Horn

In God We Trust, All Others Bring Data: Assessing the State of Outcomes Measurement for Faith-Based and Community-Based Programming.........................................179

Robert L. Fischer

Faith-Based Institutions as Project Implementers: An Innovative Economic Empowerment Intervention for Care and Support of AIDS-Orphaned and Vulnerable Children in Rural Uganda .........................................................................................213

Fred M. Ssewamala and Leyla Ismayilova

Evaluating the Potential of Faith-Based Correctional Models: A Case Study of Florida’s Faith-Based and Character-Based Institutions .............................................................237

Diana Brazzell and Nancy La Vigne

Constitutional Change and Responsibilities of Governance Pertaining to the Faith-Based and Community Initiative .................................................................................................263

Ira C. Lupu and Robert W. Tuttle

About the Contributors ................................................................................................................283

Appendix: The FBCI Research Compendium Paper Selection Process......................................285

INTRODUCTION

Community-initiated and faith-motivated efforts to meet human services needs have been a central element of the American landscape since the founding of the country (Olasky, 2008). Especially in the past several decades, faith-based and community organizations (FBCOs) have played a vital role in identifying social problems, bringing them to public attention, and providing services to the disadvantaged (Ebaugh, Chaftez, & Pipes, 2005). At times, FBCOs have been the social institutions most responsive to the needs of residents at the local level and the most viable partners for collaborative community social service delivery (Small, 2002). During the late 1990s, federal and state policy makers began to recognize the potential of partnering with smaller FBCOs because of their unique ability to respond to local needs and win the trust of their communities. This shift is exemplified by the advent of the Charitable Choice provisions of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which allowed faith-based organizations to receive federal funding for social service programs without having to change their religious identity (Small, 2002). This effort served to encourage and codify what had previously been a legally and administratively ambiguous interrelationship.

By their very nature, community-based organizations are embedded in neighborhoods and help to build trust among community members. Similarly, faith-based organizations provide social service programs as a means to serve their communities. These and other characteristics can make FBCOs uniquely suited to support individuals and families facing devastating problems, such as substance abuse, domestic violence, HIV/AIDS, crime, poverty, natural disasters, and inadequate housing. However, research and experience have pointed out that without a funding source, trained and experienced staff, and strengthened technical capabilities, FBCOs may not have adequate capacity to provide professional-level services to their respective communities.

Recognizing the important role that FBCOs play in serving communities and those in need, on January 29, 2001, President George W. Bush signed two executive orders that established the Faith-Based and Community Initiative (FBCI). One executive order created the White House Office of Faith-Based and Community Initiatives (White House, 2001b). A second executive order established the Centers for Faith-Based and Community Initiatives (CFBCI) within five federal agencies and charged these offices to “coordinate a national effort to expand opportunities for faith-based and other community organizations and strengthen their capacity to

1

better meet social needs in America’s communities” (White House, 2001a). As the FBCI progressed, three more executive orders created additional Centers for Faith-Based and Community Initiatives, resulting in centers being established in a total of 11 major federal agencies—the Department of Education, Department of Health and Human Services, Department of Housing and Urban Development, Department of Justice, Department of Labor, Department of Agriculture, Department of Commerce, Department of Veterans Affairs, Department of Homeland Security, the Agency for International Development, and the Small Business Administration (White House, 2002, 2004, 2006).

The FBCI is designed to encourage federal and state government to consider opportunities to involve FBCOs in meeting human services needs and to build the capacity of FBCOs to help meet those needs within their communities. The FBCI is grounded in the idea that governments can recognize, fund, and equip FBCOs to do their best to compassionately serve those in need.

Since the FBCI began, there have been significant shifts in the ways the federal government interacts with the faith-based and community-based nonprofit sectors in the delivery of social services. In many service areas, the FBCI has facilitated the funding of grants to FBCOs that may have not received federal funding before either because of perceived or actual barriers or because these organizations may have lacked the capacity to compete for funding. The FBCI also represents a shift in how the government interacts with the nonprofit sector by encouraging federal and state program officials to consider how they can prepare future partners to participate in the delivery of social services.

While there is some initial research on the partnerships that began in the mid- to late 1990s, much less is known about more recent developments since the advent of the FBCI in 2001. The White House Office of Faith-Based and Community Initiatives has identified 10 signature governmental innovations since the inception of the FBCI:

▪ identifying and removing unwarranted barriers that inhibit government partnerships with faith-based and community organizations

▪ expanding partnerships with community organizations through vouchers, mini-grants, and intermediary grantees

▪ implementing the FBCI through federal agencies that administer human service programs so that new and strengthened partnerships can further each agency’s mission

▪ building mutually reinforcing clusters of services so interrelated social ills can be resolved through comprehensive approaches

▪ applying a massive-scale response to the prevention, care, and treatment of those afflicted with HIV/AIDS

2

▪ expanding key elements of the FBCI in all 50 states

▪ building the capacity of nonprofit-sector leaders through training and technology

▪ expanding public-private partnerships

▪ forging partnerships with domestic and international volunteer efforts

▪ utilizing the FBCI to promote successful service models that involve FBCOs.

The innovations resulting from the implementation of the FBCI have not only led to changes in governing strategies, but have changed the landscape of social service delivery. Many of the perceived barriers that prevented FBCOs from being active participants in social services activities have been challenged, and some of these organizations have become viable partners in social service delivery networks (WHOFBCI, 2008).

Overall, the FBCI has resulted in more guidance to organizations on the appropriate role of religious activities and religious character for organizations receiving public funding. By working internally with federal as well as state and local governments, the FBCI has increased the depth and specificity of guidance over time by taking the lead in helping to standardize, clarify, and disseminate legal and regulatory guidance that initially was provided on an as-needed basis. In publishing regulations and providing education in numerous venues across the country, the FBCI has helped increase the knowledge level about the appropriate role of religious activities as well as the allowance for faith-based organizations to maintain their religious character.

THE NEED FOR RESEARCH AND EVALUATION OF THE FBCI

The underpinnings of the FBCI have broad appeal for policy makers and practitioners. However, more information is needed to better understand program implementation, best practices and challenges, and effective program components. The evaluation of faith-based and community-based programs and services is of paramount importance if the impact of the FBCI is to be understood and if faith-based organizations are to be fully integrated within a systematic framework of social service delivery (Zanis & Cnaan, 2006).

Several challenges common to evaluative research exist in investigating questions of effectiveness for FBCOs. For example, many of these organizations are small and have limited capacity for data collection. Among faith-based programs, there is often a lack of understanding of the role of faith in these programs, including whether and how faith may impact program outcomes (Fischer, 2004). In the extant faith-based outcomes research, the role of faith in faith-

3

based services tends to be viewed as a contextual factor rather than as a specific program component that directly influences clients’ experiences (Ferguson, Wu, Spruijt-Metz, & Dryness, 2007). Also, it is imperative to conceptually and operationally define what is meant by faith-based and community-based organizations if an exploration of the effectiveness of these programs is going to be undertaken. One significant challenge in the design of comparative research studies is specifying a comparison group. There can be difficulties in undertaking random assignment at the organizational level and a risk of high attrition at the participant level. The organizational mission of many FBCOs is to offer services to anyone in need (Fagan, Horn, Edwards, Woods, & Caprara 2007), thus making it ethically unacceptable to refuse treatment to some applicants as required under a random assignment approach. Also, there may be higher attrition rates for those with lower levels of faith who are randomly assigned to a faith-intensive program (Fisher & Stelter, 2006).

Researchers have begun to chronicle the characteristics that appear to differentiate faith-based organizations from community-based organizations (Wuthnow, 2004). Faith-based organizations demonstrate by their policies, practices, or mission statements that they are motivated or guided by religious ideology or that they are directly connected with an organized faith community. Other characteristics that identify an organization as faith-based are the receipt of substantial support from a religious organization or the initiation by a religious institution (Cnaan & Milofsky, 1997; Wuthnow, 2000). In contrast, community-based organizations tend to have a governing structure and staff that involve members of the community. These organizations may be less likely to have a specific association with a religious organization or ideology. Both types of organizations tend to focus more on providing emergency services and less on the organizational capacity that is often required by major funding organizations, such as the government.

As increasing numbers of grants have been awarded to FBCOs and program implementation has progressed, several evaluations and research studies of these initiatives have been launched by universities and research organizations and funded by private organizations, foundations, and the government. Currently, there are several federally sponsored research evaluations that span a range of agencies and program areas. Some examples include the Serious and Violent Offender Reentry Initiative and Comprehensive Anti-Gang Initiative funded through the Department of Justice; the Homeless Veterans Reintegration Program and the Latino Coalition Intermediary Grant Program Evaluation funded through the Department of Labor; and the Mentoring Children of Prisoners and Intermediary Model Benchmarking Study funded through the Department of

4

Health and Human Services. Many of the large-scale evaluation studies are ongoing and it will take time to report findings that measure program effectiveness.

In addition to the government-sponsored evaluations of federally funded faith-based programs, there are several small-scale research projects that have been undertaken in the field. These research projects focused on the efforts of specific churches or locally funded community initiatives (Wood, 1997) as well as thought-provoking collaborative articles by researchers and clergy who discuss their program experiences (Boddie & Cnaan, 2001) and why they do or do not apply for funding (Pipes & Ebaugh, 2002). This emerging literature is critical to exploring the nuances of program delivery in-depth. For example, how does service delivery vary by different religious traditions, and how do these programs measure “faith” per se? Also, it will be important to understand the reasons why smaller organizations apply or do not apply for government funding.

A MULTIDISCIPLINARY CONFERENCE FOR UNDERSTANDING THE FBCI

The White House Office of Faith-Based and Community Initiatives is sponsoring a national conference in June 2008 to provide a forum for the dissemination of the emerging literature and the wide range of research related to the FBCI. As part of this effort, two national calls for paper abstracts were issued that focused on research studies sponsored by the federal government and private or nonprofit organizations that examine federal and state faith-based and community-based initiatives. Using an objective scoring system, an expert panel of reviewers selected abstracts to be developed into full papers for presentation at the White House conference and inclusion in a research compendium. For more information about the calls for paper abstracts, the expert panel, and the paper selection process, see the Appendix.

Presenting information on FBCI efforts and accomplishments through a national conference and commissioned empirical research provides policy makers, researchers, and social service practitioners with much-needed information to inform decision making at all levels of government as the FBCI continues to evolve as an innovative governing strategy.

Research Studies

The studies presented in this compendium, representing the current state of legal and social science research pertaining to the FBCI, address a range of topics and programs. Because the FBCI emphasizes the building of collaborations with organizations that may have limited

5

experience with research and data collection, the development of the FBCI evidence base has evolved from descriptive studies that illustrate the various services, organizations, and program components of the FBCI to studies that focus on participant outcomes and program effectiveness.

Included in this research compendium are four types of studies that characterize the evolution of the research in this area: (1) descriptive studies of FBCI innovations, such as intermediary models, technical assistance, capacity building, and effective public-private collaborations; (2) descriptive studies examining FBCO service components, funding, accessibility to clients, barriers, and the faith orientation of organizations and clients; (3) studies focusing on participant outcomes—in some cases using comparative research designs to examine program effectiveness; and (4) an analysis of the legal and regulatory issues that govern the FBCI and influence the social service environment on which the research presented in this compendium is focused.

FBCI Innovations in Governance

Some of the research papers in this volume present studies about the innovations in governance encouraged by the FBCI and note successful partnerships as well as some of the challenges to be addressed as the FBCI develops in future years. Byron Johnson makes the case that a comprehensive approach to addressing social problems, if brought to scale, would require new partnerships and a significant influx of volunteers, many of whom could be drawn from religious congregations. Johnson examines the case of prisoner reentry, one of the signature federal initiatives of the FBCI, and assesses what it would take to bring this effort to full scale. Because religious activities can play a positive role in the lives of prisoners while they are incarcerated, and research shows that religiosity is associated with reducing negative outcomes and promoting prosocial behavior, Johnson asserts that faith-based organizations can play an important role in helping reduce recidivism. Johnson concludes that intermediaries are necessary to bring a comprehensive prisoner reentry effort to scale because these organizations serve as the bridge between ex-prisoners and the many social service providers and various governmental agencies in the areas of employment, housing, education, and counseling; provide technical assistance and oversight; and offer training to strengthen faith-based and community-based organizational capacity.

Growing the role of intermediaries, developing small grants programs, and providing technical assistance to help build FBCO capacity are important goals of the FBCI. Mitchell Brown focuses her research on organizational capacity and technical assistance issues of FBCO domestic violence service providers who participated in the evaluation of the Rural Domestic Violence and Child Victimization Enforcement Grant Program Special Initiative: Faith-based and

6

Community Organization Pilot Program, or Rural Pilot Program (RPP), funded by the Department of Justice. The RPP is implemented by intermediary organizations to administer competitive small, one-year, grants programs with FBCOs in rural areas for the provision of domestic violence services.

The research reported in this paper is drawn from a survey of funded and nonfunded organizations as well as qualitative interviews with staff members in intermediary and grantee organizations. Overall, the findings indicate that while all of the applicant organizations had many capacity needs, such as reporting requirements and staff training, the most pressing needs were not the ones anticipated, such as obtaining 501(c)(3) status and establishing evaluation benchmarks; rather, these organizations needed help to build their capacity to undertake program evaluation. Overall, Brown concludes that the RPP did help FBCOs build capacity and provided useful technical assistance, and that the funded organizations realized some modest positive changes in capacity compared with those that were not funded.

Encouraging partnerships between FBCOs and government agencies to deliver services to low-income families is an important component of the FBCI. Reporting preliminary findings from a Department of Health and Human Services sponsored study of promising practices and successful partnerships between agencies administering the Temporary Assistance to Needy Families (TANF) program and FBCOs, Courtney Barthle, Rebecca Makar, and Jeannette Hercik present an overview of five TANF and FBCO partnerships, identify the common themes associated with their success, and discuss the implications for possible future collaborations in social services provision for low-income and at-risk families. Common themes relating to the fruitful collaboration between TANF and FBCOs include recruiting a steady volunteer base and providing training to volunteers, establishing a formal referral system for clients from the TANF agency to FBCOs, building program collaborations in increments to allow infrastructure to develop, targeting services to clients that organizations have a strength in serving, and strategically locating services in a place that resonates with the target population.

The Scope of FBCO Services and Programs

Important goals of the FBCI are to reduce barriers to the participation of FBCOs in the federal funding process and to increase their organizational capacities to deliver social services. Researchers are just beginning to understand the menu of services that FBCOs deliver, differences between the types of services provided by organizational type, and barriers to participation. Three papers examine services delivered by FBCOs in different domains:

7

emergency assistance, mental health, employment, housing, and substance abuse recovery programs.

Scott Allard researches program service delivery, accessibility of services, and organizational stability among faith-based organizations and secular community-based organizations. Drawing on a survey of nonprofit service providers, this study compares services, organizational characteristics, funding, and access to faith-based and secular organizations in several urban and rural communities. Notably, the survey distinguishes between faith-based organizations that integrate religious elements into service delivery, faith-based organizations that segregate religious elements from services, and secular nonprofit organizations. Allard found that both types of faith-based organizations deliver more emergency assistance services and less mental health, substance abuse, or employment-related services compared with secular nonprofit organizations. Faith-based organizations that integrate faith into service delivery and secular nonprofit organizations are more accessible to poor populations than faith-based organizations that do not integrate religious elements into service provision. Allard concludes that overall faith-based organizations have fewer resources than secular community-based organizations, and speculates that fewer resources may, in part, lead to a different menu of services compared with community-based organizations.

Richard Hula, Cynthia Jackson-Elmoore, and Laura Reese consider the nature of faith-based and secular service delivery models in the housing sector. This study reports on the preliminary results of an ongoing survey exploring the role of FBCOs operating in housing markets in inner cities in Michigan. The authors found that, for the most part, faith-based organizations and secular community-based organizations engage in similar activities and report comparable barriers to growth, suggesting that faith-based organizations are at least as productive as secular community-based housing providers in terms of their activities and involvement in the housing arena. The authors also highlight some important differences depending on organizational type. One example is that faith-based organizations perceive local politics and regulations as more problematic to their participation in delivering services than community-based organizations. The study concludes that increasing the role of faith-based organizations in housing services may require state and local actors to help alleviate some of these participation barriers or at least the perception of barriers.

The role that faith plays in faith-based organizational settings and service components is not well understood (Wuthnow, 2004). In their pilot study, Fred De Jong and Claudia Horn examine the efforts of Gospel Relief Missions (GRMs), faith-based providers of substance abuse recovery support programs for the disenfranchised that are located throughout the country. To develop

8

measurement of the components of faith that permeate the organization and services, researchers convened work groups composed of various GRM staff members and interviewed key stakeholders about the organizational mission and the faith-based substance abuse recovery services as well as any secular services. The preliminary results indicate that clients experienced some variation in the faith-based services received, had strong faith orientations, and rated their relationships with staff highly in terms of quality and trust, which were positively associated with their spiritual growth and faith convictions.

Evaluating Participant Outcomes and Program Effectiveness

Reviewing recent published quantitative studies on the effectiveness of services delivered by faith-based and community-based organizations, Robert Fischer presents the status of the evidence base on the effectiveness of FBCO programs and the implications for growing and strengthening research in this area. Because of the relative youth of the FBCO field, much of the research is descriptive in nature and lacks rigorous methodological program evaluation designs, although Fischer notes that both the number and the quality of studies that are able to access program effectiveness are increasing.

To assess the state of the research in the field, Fischer casts a broad net and undertakes a preliminary quantitative synthesis that yields 18 studies in the realm of prisoner rehabilitation, welfare assistance, and substance abuse treatment. These studies all use comparative research designs in which program participants’ outcomes in FBCO-provided services are compared with a similarly situated group that received the usual services, secular programming, or who did not receive services. The findings of these studies show that while the overall effect of FBCO programs tends to produce somewhat better outcomes, they are of modest size. Fischer asserts that more research with larger sample sizes is needed to confirm these positive findings, but also to examine the effects for defined target populations.

Using a novel research approach that includes mixed methods and a comparative research design, Fred Ssewamala and Leyla Ismayilova present a study of an asset-building and educational intervention that is implemented by a faith-based organization to serve families caring for orphans and vulnerable children in rural Uganda. The intervention provides families in randomly selected schools with Children’s Development Accounts, which are matched savings accounts that can be used for future education or business development. The intervention also includes skill-building components, including educational workshops and monthly mentoring meetings. The findings from the group randomized trial suggest that the intervention improves children’s expectations about future careers and increases their motivation to make more careful

9

choices regarding sexual risk-taking in the treatment group compared with the comparison group. The results also suggest that poor families caring for orphans and vulnerable children in Uganda can save for the children’s educational needs and that the faith-based provider was effective in implementing the program.

Diana Brazzell and Nancy LaVigne present findings from a unique process and outcome evaluation of the Faith and Character-based Institutions Initiative (FCBI) at two Florida correctional facilities. The study findings are based on an analysis of Florida Department of Corrections longitudinal inmate data; interviews with FCBI management, staff, and volunteers; focus groups with inmates participating in the FCBI programs; and firsthand program observation. The authors used a quasi-experimental evaluation design that included a matched-comparison group. The quantitative analysis showed no evidence that the FCBI program reduced recidivism, as measured by reincarceration, for either men or women. However, FCBI staff, volunteers, and inmates endorsed the program and suggested replicating it at other facilities. The authors recommend that policy makers and corrections officials clearly define the intended outcomes of the initiative, create benchmarks for measuring success, and implement systems for tracking and analyzing outcomes data.

Legal and Regulatory Framework of the FBCI

The last paper in this volume presents an overview of the regulatory and legal framework guiding the social service environment in which this research has taken place. Ira Lupu and Robert Tuttle provide a description of the regulations that established the FBCI and the key points in the development of the relevant constitutional law. Their analysis demonstrates that the legal environment pertinent to the FBCI is complex and still developing. Against this backdrop of the changing legal environment and the inherent difficulties in translating regulations and case law into day-to-day operations, the authors note the successes and challenges of integrating the FBCI regulations into this complex legal environment. They conclude with the implications for federal and state regulations and guidance designed to facilitate the FBCI while maintaining constitutional bounds.

SUMMARY OF THE RESEARCH FINDINGS: SUCCESSES AND CHALLENGES

Although the research papers included in this compendium examine a number of social programs and use a range of research methods, it is possible to identify some broad themes about the state of the FBCO research literature.

10

There are several examples of successful FBCI innovations. These include using intermediary models, building several dimensions of organizational capacity, and providing technical assistance in the areas of prisoner reentry, TANF, domestic violence services, asset building, and HIV/AIDS prevention. Addressing a range of social problems, FBCOs provide assistance in many areas, sometimes in small pockets, and other times as part of larger programs and initiatives.

Faith-based organizations tend to provide more emergency services than community-based organizations, in part because of more limited resources; however, the research presented here highlights the diversity of needs being served by both faith-based and secular organizations. Clients in need are likely to have multiple problems that can be addressed more effectively in a holistic manner and can benefit from an individualized approach to providing assistance and skills. These organizations also may be equipped to address these issues in ways that accommodate clients’ faith and culture and recognize family and community contexts. Thus, it is important for stakeholders to consider comprehensive and scaleable strategies.

Some of the papers in this compendium provide examples of intermediaries that effectively build FBCO capacity. Notably, using mini-grant programs among domestic violence service providers increased organizational capacity among organizations that were funded by the grants program; and, interestingly, capacity also increased among organizations that intermediaries worked with but were not funded by the grants program. This finding suggests that intermediaries can assist organizations in more ways than just providing access to funding. There also is evidence of the successful use of intermediaries in prisoner reentry programs that build organizational capacity and facilitate service delivery by leveraging a broad base of volunteers from congregations and community organizations to serve as mentors.

In addition, studying successful partnerships between TANF agencies and FBCOs has shown that FBCOs can provide individualized support services and mentoring to families to help achieve successful outcomes. An important finding is that faith-based providers and secular community-based organizations in urban and rural areas serve predominantly low-income populations. And this research emphasizes the importance of the location of FBCOs in terms of proximity to and ease of access for clients, but also in fostering comfort levels and trust.

Further, these studies point to the improving capacity of local faith-based and community-based providers and government agencies to undertake data collection and implementation of program interventions that include an extensive research component. For instance, the pilot research undertaken with the Gospel Rescue Missions (GRM), a large network of faith-based providers

11

that had not previously participated in any major research project, allowed researchers to “open the black box of services” and attend meetings and interview staff to understand the organizations’ and clients’ faith orientations. GRM employees benefited from training in Web-based survey assessments and the collection of participants’ outcome data.

Similarly, the partnership highlighted between a faith-based organization and a university to implement a program intervention in rural Uganda, which included randomization of treatment at the school level, shows that research on these types of activities is starting to take root in other countries as well. FBCOs can act as a full research partner in these efforts.

State and local government agencies are also starting to collect more data and give researchers access to facilities, as demonstrated by the Florida Department of Corrections allowing researchers to analyze program participation data and to conduct participant interviews and observation within the prisons.

The FBCI promotes the provision of social services by both faith-based and community-based programs. An increasing number of research studies are using comparative research designs to examine whether there are differential effects of faith-based programs compared with secular programs, usual services, or no services. Several papers highlight the effects of FBCO programs on participants’ behavioral outcomes. The assessment of the most recent 18 studies that use comparative designs shows a mix of negative, null, and positive findings about the effects of faith-based services on outcomes. However, pooling the results of these studies, the mean effect of faith-based programs is statistically significant and positive, although small in magnitude. Pooling the studies by topic shows a statistically significant positive effect for welfare services that is of moderate effect size, and a smaller, though still positive effect for prisoner reentry programs.

One of the studies in this compendium found a positive effect of a program on reducing reincarceration, although it was not statistically significant. Another study found a significant treatment effect that included increasing educational aspirations and reducing risk behaviors. Although the findings in these two studies varied in their statistical significance, the program staff and clients in both programs found them to be worthwhile. While these studies represent some of the most rigorous designs in the literature, there are limitations to the findings because participants were not fully randomized into treatment and control groups, which can introduce bias into the estimates of program effects due to selection issues.

12

While there are clearly notable successes, the empirical research presented also points to challenges in program implementation and evaluation. One challenge is that while there are examples where organizational capacity increased, overall faith-based organizations generally have lower levels of organizational capacity than community-based organizations. This makes it difficult to offer a wide range of programs, hire staff, provide training to staff and volunteers, and participate in labor-intensive research projects. In addition, sometimes FBCO staff assume that they have the capacity to undertake research, but in reality they may not have a clear understanding of program evaluation, survey assessment, and data collection. One study noted that faith-based organizations also may in fact not have organizational barriers to participation to funding and service delivery, but may perceive that local political barriers are present. Moreover, the evolving case law and regulations about allowable activities for grantees receiving federal funding can be difficult to translate into the everyday practices of FBCOs, which may make these organizations initially wary of participating in or conducting research. While a significant amount of effort and progress has been made to increase the capacity of FBCOs to undertake research, and ongoing technical assistance and guidance has been provided that stresses the importance of research to potential and actual grantees, there are some important next steps to help guide ongoing research on the FBCI.

NEXT STEPS

Based on the findings of this important initial research on the FBCI, there are a number of steps that could be taken to move the field and the FBCI forward.

Examining FBCO Services

First, it will be important to include specific examination of FBCO services and programs in government and academic research on social service delivery. Evaluation and research have become increasingly important for federal programs as well as for the social service delivery network as a whole. Policy makers and practitioners want to know the results and outcomes achieved from investments in this sector. Evaluation and outcomes research presents the opportunity to determine the best ways to leverage the unique strengths of FBCOs in serving particular populations through different kinds of service delivery approaches.

Conducting Mixed-Method Research

As the burgeoning research on FBCOs evolves, there is a need to move beyond descriptive research and short-term outcomes to a mixed-method (i.e., qualitative and quantitative) approach that provides a more comprehensive understanding of program outcomes over time. Incorporating the multiple perspectives, for example, of participants, program staff, and

13

community members on program implementation can help to elucidate the impact of the program on those involved in the work of FBCOs. This broad data collection strategy will improve understanding of the context in which FBCOs provide services. By using longitudinal studies that include comparison groups in particular, researchers can explore the sustained effects of the FBCI on client outcomes over time and begin to uncover the specific aspects that contribute to successful social service delivery among FBCOs.

Considering the Effects of Policy and Practice

As the field gains a better understanding of the role of FBCOs in service delivery, research could consider the effects of federal, state, and local government policies and practices on their operations. Research presented in this compendium raises questions about the degree to which the social service sector understands the regulatory and legal framework that has developed around the FBCI. Research could shed more light on the context in which FBCOs operate by determining grantees’ and potential grantees’ levels of understanding and areas of confusion about the regulatory framework implemented over the past 7 years. Having a better understanding of grantees’ perspectives could help policy makers determine the need for further clarification.

Standardizing Definitions of FBCOs

The research field could help address some of the definitional issues that can cause confusion in the service delivery arena. Specifically, the lack of an agreed-on definition of what constitutes a faith-based organization or a community-based organization makes it difficult to isolate how these organizations operate and succeed as compared with other organizations. A first step in addressing this issue is to obtain consensus on a standardized definition of faith-based and community-based organizations.

Specifying the Faith Component

Another focus area for future research centers on the need to specify the faith component in faith-based programs and services. Inasmuch as many faith-based organizations conceptualize faith as a primary or central component in their programs and services, future research needs to explicitly address this construct in evaluation studies.

Exploring the Community Factor

Lastly, as reflected in this compendium, a considerable amount of recent research has focused particularly on faith-based organizations. However, one study presented here suggests that smaller community-based organizations can play a prominent role in social service delivery.

14

Thus, one important area for future FBCI research would be to explore what could be termed the community factor or the extent to which both faith-based and community-based organizations are embedded in the broader communities they serve. Examining these broader organizational relationships as well as the individual relationships with clients will provide insight into the operational processes of faith-based and community-based organizations. This research avenue will also enhance understanding of the level of organizational acceptance within communities. Understanding this broader community context is also essential to identifying and addressing participatory barriers for faith-based and community-based organizations. Examining the complexity of these organizational relationships will enhance understanding of the FBCI overall as well as the implications for and impact on the lives of the individuals, families, and communities served.

REFERENCES

Boddie, S. C. & Cnaan, R. (2001). Philadelphia Census of Congregations and their involvement in social service delivery. Social Service Review, 75, 559–580.

Cnaan, R., & Milofsky, C. (1997). Small religious nonprofits: A neglected topic. Nonprofit and Voluntary Sector Quarterly, 26(4), S3–S13.

Ebaugh, H. R., Chafetz, J. S., & Pipes, P. (2005). Funding good works: Funding sources of faith-based social service coalitions. Nonprofit and Voluntary Sector Quarterly, 34(4), 448–472.

Fagan, P., Horn, C., Edwards, C. W., Woods, K. M., & Caprara, C. (2007). Outcome based evaluation: Faith-based social service organizations and stewardship. Special Report #13. Washington, DC: The Heritage Foundation.

Ferguson, K., Wu, Q., Spruijt-Metz, D., & Dyrness, G. (2007). Outcomes evaluation in faith-based social services: Are we evaluating faith accurately? Research on Social Work Practice, 17(2), 264–276.

Fischer, R. L. (2004). The devil is in the details: Implementing outcome measurement in faith-based organizations. Nonprofit Management & Leadership, 15(1), 25–40.

Fischer, R. L., & Stelter, J. D. (2006). Testing faith: Improving the evidence base on faith-based human services. Journal of Religion & Spirituality in Social Work, 25(3/4), 105–122.

Olasky, M. (2008). The tragedy of American compassion. Wheaton, IL: Crossway Books.

Pipes, P., & Ebaugh, H. R. (2002). Faith-based coalitions, social services, and government funding. Sociology of Religion, 63(1), 49–68

15

Small, M. (2002). Achieving community justice through faith-based initiatives. Behavioral Sciences and the Law, 20, 411–421.

White House. (2001a, January 29). Executive Order 13198. Agency Responsibilities with Respect to Faith-Based and Community Initiatives. Retrieved from http://www.whitehouse.gov/news/releases/2001/01/20010129-3.html

White House. (2001b, January 29). Executive Order 13199. Establishment of White House Office of Faith-Based and Community Initiatives. Retrieved from http://www.whitehouse.gov/news/releases/2001/01/20010129-2.html

White House. (2002, December 12). Executive Order 13280. Responsibilities of the Department of Agriculture and the Agency for International Development with Respect to Faith-based and Community Initiatives. Retrieved from http://www.whitehouse.gov/news/releases/2002/12/20021212-5.html

White House. (2004, June 1) Executive Order 13342. Responsibilities of the Department of Commerce and Veterans Affairs and the Small Business Administration with Respect to Faith-Based and Community Initiatives. Retrieved from http://www.whitehouse.gov/news/releases/2004/06/20040601-1.html

White House. (2006, March 7). Executive Order 13397. Responsibilities of the Department of Homeland Security with Respect to Faith-Based and Community Initiatives. Retrieved from http://www.whitehouse.gov/news/releases/2006/03/20060307-5.html

White House Office of Faith-Based and Community Initiatives. (2008). The quiet revolution: The President's Faith-Based and Community Initiative: A seven-year progress report. Washington, DC: Author.

Wood, B. A. (1997). First African American Episcopal Church and its social intervention in South Central Los Angeles. Unpublished doctoral dissertation, University of Southern California, Los Angeles.

Wuthnow, R. (2000). How religious groups promote forgiving: A national study. Journal for the Scientific Study of Religion, 39(2), 125–139.

Wuthnow, R. (2004). Saving America? Faith-based services and the future of civil society. Princeton, NJ: Princeton University Press.

Zanis, D. A., & Cnaan, R. A. (2006). Social service research and religion: Thoughts about how to measure intervention-based impact. Journal of Religion & Spirituality in Social Work, 25(3/4), 83-104.

16

NOT BY GOVERNMENT NOR FAITH ALONE: RETHINKING PRISONER REENTRY

Byron Johnson

The process of reintegrating ex-prisoners into society has been a persistent and challenging social issue for decades. What is different in the contemporary landscape, however, is the sheer number of prisoners returning to American communities each year (Osborne & Solomon, 2006). And it is becoming clear that this unprecedented development is not a temporary trend. Between 1980 and 2006, the U.S. prison population increased by 467% (from 319,598 to 1,492,973) and the parole population increased by 362% (from 220,438 to 798,202) (U.S. Department of Justice, 2006). The increase in the number of prisoners returning to communities across the country has fostered a national debate about how best to handle what has become known as the prisoner reentry crisis and one of the most challenging dilemmas in U.S. correctional history (Travis, 2005).

A number of well-known correctional programs have been implemented over the years to help manage the difficult adjustment period when prisoners transition back to society. Halfway houses, community corrections, intensive supervision, and community reintegration programs represent a few of the various postrelease efforts designed to facilitate prisoner reentry to society, while simultaneously ensuring public safety (Petersilia, 2003). But despite corrections expenditures now in excess of $60 billion annually, the likelihood of a former prisoner succeeding in the community has not improved (Bauer, 2002). Indeed, about two thirds of all offenders released from prison are rearrested within 3 years of their release (Langan & Levin, 2002). In addition, growing parole caseloads have made effective case management by parole officers an increasingly difficult proposition, with an occupational by-product of increased stress on parole officers—a problem having to be addressed by a number of parole agencies (Finn & Kuck, 2003). As a result of the difficulty in effectively managing large numbers of returning prisoners, there is increasing concern that the number of ex-prisoners returning to society could pose a threat to public safety.

Although the problems confronted by ex-prisoners returning to society are readily identifiable, public efforts to address reentry and aftercare problems have been limited in scope (Travis & Vischer, 2005). In general, policy makers have been reluctant to support correctional policies that endorse or appear to favor offender treatment, job training, and counseling for ex-prisoners and their families in the community. Such efforts can easily be interpreted as taking a “soft on

17

crime” approach. One might argue that a prisoner reentry plan including such programs has the potential to significantly reduce recidivism and thus improve public safety. However, few policy makers have been willing to publicly defend such programs. Not surprisingly, law-and-order crime policies have consistently trumped those favoring offender treatment models (Cullen, 2002).

Moreover, the lack of a comprehensive governmental response to prisoner reentry is influenced by budgetary constraints. Creating new offender treatment and support programs in prisons as well as in communities would place a significant financial burden on correctional budgets that are already regarded by many as too high. In an era of finite resources and ever-tightening budgets, efforts to significantly expand existing educational, vocational, and counseling programs in prisons and communities have not received serious consideration. However, it is both unrealistic and unwarranted to lay the sole responsibility of comprehensive prisoner reentry on government.

Although often overlooked, the role of religion, religious volunteers, religious programs, and faith-based organizations in the criminal justice system has been a constant in U.S. history. This oversight is unfortunate because numerous theoretical perspectives, published research, and common sense suggest that communities of faith have the potential to be a powerful partner with government in the development of a comprehensive prisoner reentry plan.

The following section reviews research documenting the role of religion in prisons and prisoner reentry, and reviews research connecting religion to crime reduction as well as prosocial behavior; and thus provides a basis for inclusion of a faith-based approach to prisoner reentry.

THE RELEVANCE OF RELIGION IN PRISONS AND PRISONER REENTRY

The Role of Religion in Prisons

The evolution of the U.S. correctional system has been accompanied by the continual influence of religion and religious workers. For example, terms such as corrections, penitentiary, reformation, restoration, and solitary confinement can be traced to religious origins (McGowen, 1995; Peters, 1995).

The historical role of religion in prisons continues to be prominent and pervasive in correctional institutions today. Faith-motivated volunteers in prisons are as likely to be involved in life-skills training or instruction in General Equivalency Diploma (GED) programs as they are to conduct

18

Bible studies or lead worship services. Thus, religious volunteers have played and continue to play a vital role in the vast majority of U.S. correctional institutions.1 Indeed, beyond work, education, or vocational training, religious activities attract more participants than any other type of personal enhancement program offered inside a prison.

There are many ways that religion can be consequential for prisoners and ex-prisoners. However, where correctional decision-makers and policy stakeholders are concerned, the one overriding outcome is whether an intervention reduces recidivism. In the mid-1990s, Prison Fellowship (PF), a nonprofit religious ministry to prisoners, commissioned research to determine the effects of faith-based interventions on prisoner recidivism. Utilizing a quasi-experimental design, the study examined the influence of religious programs on prisoner adjustment (i.e., institutional infractions or rule violations) and recidivism rates (i.e., postrelease arrests) in two matched groups of inmates from four adult prisons in New York State.2 One group had participated in programs sponsored by PF; the second group had no involvement with PF programs. Researchers found that after controlling for level of involvement in PF-sponsored programs, inmates who were most active in Bible studies were significantly less likely than their matched comparison counterparts not participating in the PF-sponsored programs to be arrested during the one-year follow-up period (Johnson, Larson, & Pitts, 1997).

A second study, conducted with an additional 7 years of follow-up data, documented that after dividing the sample into groups of high and low levels of participation in Bible studies, high-level participants were less likely to be rearrested at 2 and 3 years postrelease (Johnson, 2004). The study concluded that more research is necessary to determine how religion might be related to offender rehabilitation, inmate adjustment, and prisoner reentry. This small but growing body of research suggests that participation in religious programs and activities can contribute to positive inmate adjustment while in prison, as well as reduce the likelihood of recidivism following release from prison (Johnson, 2003, 2004).

From Bible Study to Faith-Based Prison Programs

An overarching implication of this relatively new body of research is that religious volunteers and faith-based programs have the potential to play a significant role in contributing to prison management, safety, and offender rehabilitation. For example, preliminary research suggests that faith-based dorms and housing units have the potential to significantly counteract the negative and often debilitating prison culture that permeates so many U.S. correctional institutions (Clear & Sumter, 2002; Johnson, 2003).

19

A 6-year evaluation of a faith-based prison program called the InnerChange Freedom Initiative (IFI)3 found that inmates completing the program were significantly less likely than a matched group of offenders4 to be rearrested (17% versus 35%) or reincarcerated (8% versus 20%) during a 2-year follow-up period (Johnson, 2003). The study revealed a stark contrast between the areas of the prison controlled by the faith-based program as compared with those areas housing prisoners from the general population, which were typified by the presence of a distinct prison code of behavior that often condones rule-breaking and other inappropriate behaviors.

In contrast, the faith-based side of the prison was characterized by educational classes, study, work, worship services, little free time, and the absence of television. Further, the faith-based program enjoyed an atmosphere promoting forgiveness, honesty, and personal accountability. Faith-based efforts like IFI and Kairos, another faith-based prison program, are designed to discourage antisocial and destructive behavior and to encourage transparency, contrition, and spiritual transformation—all of which runs counter to the pervasive prison code or culture that often works to undermine the very premises on which a rehabilitation model is based (Clemmer, 1958).5

Preliminary research lends support for the notion that faith-based units can create an environment that is conducive to effective treatment and to rehabilitation programs more generally (Johnson, 2003). In this way, faith-based interventions have the potential to enhance the achievement of a secular goal and civic good; namely, lower recidivism.

FAITH-BASED PRISONER REENTRY: STRENGTHS AND SHORTCOMINGS

As important as volunteer work within correctional facilities might be, it does not diminish the fact that reentry and aftercare tend to be largely overlooked by most religious volunteers and organizations. Compared with reentry, prison ministry is a much easier task to pursue and a safer service opportunity in what many consider to be an unsafe environment. Prisoners often appreciate the attention they receive from the outside world, and these exchanges tend to be overwhelmingly positive and nonthreatening for volunteers. Prison ministry, therefore, can be found in many U.S. congregations, and thousands of religious volunteers visit prisons every day.

Similarly, faith-based organizations disproportionately opt for in-prison ministry as opposed to out-of-prison services because reentry and aftercare are anything but easy or safe. For example, Prison Fellowship Ministries, the largest faith-based prison ministry in the United States, has always recognized that reentry and aftercare are vitally important, but their efforts have only

20

been marginally involved in reentry and aftercare, an oversight recently acknowledged by the organization. PF has stated an intention to remedy this imbalance by significantly expanding its emphasis on aftercare.6

While the disproportionate emphasis of volunteerism is in prisons rather than on aftercare in communities, it would be inaccurate to suggest that faith-based prisoner reentry programs are nonexistent. Unfortunately, it is unclear how many faith-based reentry programs are operational, although it is likely they exist in many of the communities where prisons are located. These reentry programs tend to be small, isolated, and in need of coordination as well as evaluation.

The Link Between Religion and Crime Reduction

Systematic reviews and one meta-analysis of religion and crime literature have provided evidence that religious commitment and involvement are linked to reductions in delinquent behavior and deviant activities (Baier & Wright, 2001).7 Recent evidence suggests that such effects persist even if there is not a strong prevailing social control against delinquent behavior in the surrounding community. For example, several studies found that young Black males from poverty tracts in Boston, Chicago, and Philadelphia were much less likely to be involved in crime and delinquent behavior if they regularly attended church (Freeman, 1986; Johnson, Larson, Jang, & Li, 2000a). Similarly, research has found that highly religious low-income youths from high-crime areas are less likely to use drugs than less religious youths in these same disadvantaged communities. Further, these highly religious teens from crime “hot spots” were less likely to use drugs than less religious teens from middle-class suburban communities or “good places” (Jang & Johnson, 2001).

There is also evidence that religious involvement may lower the risks of a broad range of delinquent behaviors, including both minor and serious forms of criminal behavior (Evans et al., 1996). Research also shows that religious involvement may have a cumulative effect throughout adolescence and may significantly lessen the risk of later adult criminality (Johnson, Jang, Larson, & Li, 2001). Studies have shown that religion can be used as a tool to help prevent high-risk urban youths from engaging in delinquent behavior (Johnson, Larson, Jang, & Li, 2000b; Johnson et al., 2001). For example, the African-American church may play a key role in reducing crime among Black youths from urban communities (Johnson et al., 2000a). It is precisely these communities of disadvantage where many ex-prisoners will be returning.

There are many theoretical perspectives that help explain why and how religious beliefs and practices may ultimately influence behavior. To review these theories is beyond the scope of this paper; however, in sum, one can reasonably say that religious involvement helps some

21

individuals to learn prosocial behavior (i.e., actions that emphasize concern for others’ welfare). These prosocial skills may instill a greater sense of empathy toward others and thus lessen the likelihood of committing acts that harm others (Johnson et al., 2000a).

Similarly, once individuals become involved in deviant behavior, it is possible that participation in religious activities may help steer them back to a course of less deviant behavior and away from potential career criminal paths (Johnson, 2008). One important study found that religion, as indicated by religious activities, reduced the likelihood of adult criminality as measured by a broad range of criminal acts. The relationship persisted even after secular controls were added to the model. Further, the finding did not depend on social or religious contexts (Evans, Cullen, Dunaway, & Burton, 1995).

In sum, religiosity is now beginning to be acknowledged as not only a key protective factor that buffers or protects from harmful outcomes, but also as a variable promoting prosocial behavior (Johnson, 2007). If congregations can be viewed as institutions dedicated to improving the plight of at-risk populations, it may be that faith-based and community-based organizations represent key factors in helping ex-prisoners transition back to society.

HARNESSING HUMAN AND SPIRITUAL CAPITAL THROUGH INTERMEDIARIES

In January 2001, President Bush signed an executive order establishing the White House Office of Faith-Based and Community Initiatives.8 Over the next several years, Centers for Faith-Based and Community Initiatives were created in 11 federal agencies through a series of executive orders.9 In his executive orders and speeches on this initiative, the President acknowledged the long tradition of faith-based and community-based organizations helping Americans, especially those confronting serious disadvantages. President Bush also believed that the federal government had not been a very good partner to faith-based and community-based groups working to target serious social problems and that the federal government had made it difficult for faith-based and community-based groups to compete for funds on an equal standing with secular nonprofit service providers.

A 2001 White House report, An Unlevel Playing Field, systematically reviewed federal funding and identified the barriers to effective government partnerships with faith-based and community-based organizations.10 For example, the report revealed that the Office of Justice Programs at the Department of Justice estimated it would award about 0.3% of total discretionary grant funds

22

($1.9 million of $626.7 million) in fiscal year 2001 to faith-based organizations and 7.5% ($47.2 million) to community-based providers of social services.

Since 2001, considerable progress has been made in alleviating obstacles that have deterred faith-based and community-based organizations from seeking grants to build capacity and thereby strengthen outreach to underserved populations, including prisoners and ex-prisoners (prior to 2001 references to faith-based groups were virtually absent from federal funding announcements covering social service delivery or demonstration projects). In addition, conferences for faith-based and community-based groups have been offered in all regions of the United States in order to identify the federal funding processes.

President Bush has been a strong supporter of public-private partnerships whose mission it is to assist offenders, prisoners, ex-prisoners, and their children. Although he has indicated that the government has an important role to play when it comes to prisoner reentry, he has been equally clear that government is not equipped to provide the mentoring, care, and social supports that are essential for any effective and holistic plan for prisoner reentry. Stated differently, government cannot effectively address the prisoner reentry crisis by itself, nor can faith-based organizations and individuals effectively address the prisoner reentry problem by themselves. In fact, faith-based and secular partnerships represent a very viable avenue for developing an effective prisoner reentry strategy.

In their book Prisoner Reentry and Crime in America, Travis and Visher (2005) ask two important questions: “Is it possible to imagine a world in which the agencies of the justice system—corrections, police, courts, and parole—work together with other public and private institutions—housing providers, workforce development agencies, drug treatment providers, foster care agencies, and churches and other faith institutions—to systematically reduce the risk of failure around the time of reentry?...What would such a strategy look like?” (pp. 255–256).

As a result of President Bush’s belief in the role of intermediaries as well as his interest in prisoner reentry, two major prisoner reentry initiatives are now beginning to provide some preliminary answers. A third and related initiative, commonly referred to as the Second Chance Act, has recently been signed into law in early April 2008.11

Ready4Work

In 2003, the Department of Labor launched Ready4Work, a 3-year pilot program to address the needs of ex-prisoners through faith-based and community-based organizations. Ready4Work emphasized job training, job placement, case management, mentoring, and other aftercare

23

services. Faith-based and community-based organizations were selected to provide services to adult ex-offenders in 11 cities.12

Ready4Work targeted participants with a high probability of recidivism.13 Ex-prisoners in Ready4Work had extensive criminal histories and half of the participants had been arrested previously five or more times (Farley & Hackman, 2006). After individuals entered the program, they were eligible for services lasting up to one year. Participants were also matched with mentors in one-to-one and/or group mentoring relationships. Job placement specialists helped participants find employment and case managers continued to provide assistance after participants were employed.

The Ready4Work pilot ended in 2006 and the results indicate a total of 4,482 former prisoners enrolled in Ready4Work. Of these ex-prisoners, 97% received case management services, 86% received employment services, and 63% received mentoring services. Ready4Work sites placed 2,543 participants (57%) into jobs, with 63% retaining jobs for 3 consecutive months after placement (Farley & Hackman, 2006).

Public/Private Ventures (PPV)―an action-based research, public policy, and program development organization―oversees the Ready4Work demonstration project as an intermediary. PPV reported that only 2.5% of Ready4Work participants were reincarcerated within 6 months and 6.9% were reincarcerated at the one-year postrelease mark. Although Ready4Work does not incorporate a randomized design, the preliminary findings are impressive.

Over 60% of Ready4Work participants received mentoring as part of their services. Research has shown that mentoring matters for youths, but this study demonstrated that mentoring affects outcomes for Ready4Work participants. The results indicate that those Ready4Work participants who met with a mentor remained in the program longer, were twice as likely to obtain a job, and were more likely to stay employed than participants who did not meet with a mentor (Farley & McClanahan, 2007). PPV researchers concluded that “while mentoring alone is not enough, supportive relationships—which can be fostered through mentoring programs—should be considered a core component of any reentry strategy” (McClanahan, 2007, p. 3).

The Ready4Work initiative in 11 cities represents a major demonstration project. Overall, Ready4Work provides an important preliminary snapshot of what can potentially be achieved when an intermediary brings together public and private partnerships to address prisoner reentry in a comprehensive and coordinated strategy. The early results from Ready4Work support the notion that a comprehensive prisoner reentry plan is possible without a massive expansion of the

24

existing criminal justice system. Additionally, Ready4Work has helped to highlight the work of faith-based and community-based groups addressing prisoner reentry, such as Exodus Transitional Community in Harlem, Word of Hope Ministries in Milwaukee, or the Safer Foundation of Chicago.

The Prisoner Reentry Initiative

The Prisoner Reentry Initiative (PRI) was announced by President Bush in 2004 and grew out of the Ready4Work project. The PRI was designed to further test the proposition that prisoner reentry could be effectively accomplished with a comprehensive strategy designed to draw heavily from partnerships with faith-based and community-based groups. This project helps to connect former prisoners with faith-motivated groups as well as secular community-based organizations willing to help ex-prisoners locate employment and to stay out of trouble by following prosocial paths. Currently there are 30 PRI grantees across the country providing mentoring, employment, and other transitional services to thousands of ex-prisoners.

The PRI sites began serving program participants in spring 2006, and the preliminary results are promising. It is important to note, however, that these early outcomes are very preliminary and they are not based on a randomized design with strict controls. A total of 10,361 PRI participants had been enrolled as of November 2007, and about 6,000 participants have been placed into jobs. Participants’ one-year postrelease recidivism rate is currently 20%.

As can be seen in Table A-1 (see Appendix A), 9 of the 30 PRI grants were awarded to faith-based organizations; 21 grants were awarded to community-based organizations, and all but three of these secular organizations reported working with faith-based organizations. Indeed, collaborations with faith-based organizations appear to be equally important for faith-based as well as community-based PRI recipients. These alliances lend credence to the premise that faith-based and secular partnerships can be critical in establishing a social support network that is necessary for comprehensive and coordinated prisoner reentry.

A COMPREHENSIVE AND SCALABLE PRISONER REENTRY PLAN

Any prisoner reentry plan that is comprehensive and can achieve scale will require new people and programs that do not currently exist in most jurisdictions. It is unlikely that the government will or can provide these programs; thus, faith-based and community-based groups represent a critical piece of the reentry puzzle that has yet to be tapped in a systematic fashion. A sustainable comprehensive prisoner reentry plan will require that partnerships between faith-based and

25

secular as well as national and community-based groups are strengthened. A healthy atmosphere of mutual respect must replace the suspicion that too often typifies relations between public and private organizations as well as between religious and secular organizations that often share similar social service missions. This represents another way that intermediaries can serve as a strategic bridge to build alliances to more effectively address social problems.14

For example, religious individuals and faith-based groups need to recognize that ongoing training regarding correctional issues is something to be coveted rather than merely tolerated. Religious volunteers should be required to undergo basic training regarding custodial and security issues before being allowed to do volunteer work. Further, ongoing training for religious volunteers should be endorsed as well as widely promoted by faith-based organizations. This is especially true for faith-motivated volunteers who are interested in mentoring prisoners and ex-prisoners.

Faith-based groups need to understand that accountability, assessment, and evaluation of their efforts—something that will surely follow if these groups partner with government—is an extremely useful tool. Overstating program effectiveness without empirical evidence has often been problematic for many religious volunteers and faith-based organizations. If faith-based efforts are to be considered seriously, then religious volunteers need to understand that faith-based programs should be evaluated objectively. This kind of accountability will go a long way in improving relationships with other private and public groups whose confidence faith-based groups need in order to provide a comprehensive and coordinated response to prisoner reentry.

The Role of Volunteerism in Prisoner Reentry

To effectively impact prisoner reentry will require somewhat of a paradigm shift. In addition to leading a Bible study in prison, many religious volunteers will need to consider developing strategies to improve, for example, the housing and employment conditions for ex-offenders already living in the community as well as prisoners who will eventually be returning home.

Further, the importance of mentoring relationships that are established in prison and carry over to the community cannot be overemphasized. Research confirms that mentoring matters not just for kids, but also for adults. However, there is a severe shortage of mentors, especially for prisoners and ex-prisoners. This is precisely why volunteer-rich communities of faith are uniquely positioned to assist (Cornwell & Harrison, 2004; Musick & Wilson, 2007). However, faith communities have not been approached in any meaningful way on a national scale to provide these mentors. Another possibility is having properly trained volunteers to specifically assist parole and other community corrections personnel. Ultimately, a comprehensive prisoner reentry

26

plan will require a very large number of committed and trained volunteers and partners as well as a willingness to connect them, along with their varied networks of social and spiritual support, to correctional, governmental, and secular entities committed to prisoner reentry and aftercare. Without a comprehensive approach that coordinates public and private, and secular and faith-based partnerships, the lack of effective prisoner reentry support will persist.

The lack of housing, employment, transportation, counseling, and mentoring are substantial obstacles that make the transition from prison to society very difficult for ex-prisoners. Tackling these problems will require a great deal of new human and financial resources as well as the participation of key community leaders. Thus, any comprehensive strategy for confronting the problems of prisoner reentry will require an infusion of an unprecedented number of new volunteers that have or can develop strategic alliances focused on each of the problems ex-prisoners encounter.

Some organizations have developed immensely successful strategies for recruiting mentors. For example, in the past 9 years, Big Brothers and Big Sisters of America (BBBSA), the largest mentoring organization in the world, has more than doubled the number of children served.15 Along with many other grantees, BBBSA has been a significant partner in the Mentoring Children of Prisoners (MCP) efforts nationally. While it took BBBSA 95 years to make 120,000 matches on their own, there are currently 250 MCP programs in 48 states partnering with more than 6,000 churches to serve at least 100,000 children.16 The experience of Amachi, a program that recruits mentors to be matched with children of prisoners, also confirms that when people are asked to volunteer, many will do so (Musick & Wilson, 2007). A comprehensive plan for prisoner reentry that draws heavily on volunteers will need to develop strategies for recruiting mentors, although this effort will certainly look different than either the BBBSA or Amachi models.

The vast majority of the many thousands of correctional volunteers tend to come from volunteer-rich religious congregations (Musick & Wilson, 2007), and there are approximately 375,000 congregations in the United States.17 Religious congregations not only mobilize volunteer labor for the church itself, but are feeder systems for many other nonprofit and voluntary organizations. Further, religious volunteers do not necessarily choose between volunteering for the church or a secular organization; many individuals do both (Clain & Zech, 1999; Cornwell & Harrison, 2004). Surveys have consistently found a positive association between religious affiliation and attendance and charitable behavior, both in terms of financial giving and volunteering (Brooks, 2006; Independent Sector, 2001). Putnam (2001) echoed this finding when he observed that “houses of worship build and sustain more social capital—and social capital of

27

more varied forms— than any other type of institution in America. Churches, synagogues, mosques and other houses of worship provide a vibrant institutional base for civic good works and a training ground for civic entrepreneurs. Roughly speaking, nearly half of America’s stock of social capital is religious or religiously affiliated, whether measured by association memberships, philanthropy, or volunteering” (p. 54).

Although research that confirms volunteers will respond if approached with the right message, it is not enough to simply attract large numbers of volunteers. The coordination and mobilization of volunteers and organizations is equally important.

The Role of the Intermediary in Prisoner Reentry

As discussed earlier, developing a truly comprehensive prisoner reentry plan is difficult because there are so many challenges that complicate an ex-prisoner’s effort to successfully transition back to society. Focusing on housing without giving proper consideration to employment is a recipe for failure. Likewise, concentrating on transportation without giving consideration to mentoring and other social supports is likely to be unsuccessful. Any comprehensive prisoner reentry plan must coordinate a strategy to overcome all the major obstacles to successful reentry. Ready4Work and the PRI have provided a very preliminary and positive glimpse of a multifaceted reentry plan that owes much to the contribution of an intermediary organization to coordinate such efforts. In the case of Ready4Work, Public/Private Ventures is overseeing the demonstration project as a national intermediary. The PRI has utilized a different approach by essentially funding faith-based and community-based organizations to serve as local intermediaries coordinating reentry efforts (see Table A-1).

Honig (2004) offered a helpful definition of intermediaries and the role that they play: “Intermediaries are organizations that occupy the space in between at least two other parties. Intermediary organizations primarily function to mediate or to manage change in both those parties. Intermediary organizations operate independently of these two parties and provide distinct value beyond what the parties alone would be able to develop or to amass by themselves. At the same time, intermediary organizations depend on those parties to perform their essential functions” (pp. 65–87).

In recent years, the federal government has begun to use intermediaries to help faith-based and community-based organizations build capacity, strengthen programs, and improve the delivery of social services. Perhaps the best recent example is the Compassion Capital Fund (CCF), established by Congress in 2002, which provides funds to be distributed by the Department of

28

Health and Human Services to intermediary organizations across the country to provide training as well as technical and financial assistance to faith-based and community-based organizations.

The role of faith-based and community-based intermediaries in social service provision is still relatively new and underdeveloped. However, intermediary organizations may be the most important, yet underutilized, element in building successful prisoner reentry models that are intended to work with volunteers, especially volunteers who come from religious congregations. Intermediaries can be a bridge between ex-prisoners and social service providers and governmental agencies. Intermediaries can coordinate reentry efforts of faith-based and community-based organizations, volunteers, social services, mentors, and parole officers. Additionally, intermediaries can serve important roles by providing technical assistance and oversight to groups and organizations and offer ongoing training to strengthen capacity and sustainability.

Without this level of assistance, it is likely that small grassroots groups will ultimately fail. According to Mike Doyle, Executive Director of the Cornerstone Assistance Network, a faith-based intermediary organization in Fort Worth, Texas, failure to develop a sound organization will cause even successful programming to suffer if not surrender to financial and reporting pressures (M. Doyle, personal communication, March 5, 2008). Intermediaries can play a key role in coordinating the efforts of fragmented faith-based and community-based organizations (Fink & Branch, 2005). Too often these small organizations operate in relative isolation from each other and as a result are not able to build or sustain capacity. Rather than working in isolation, influential and well-networked intermediaries are well positioned to play a key role in coordinating resources locally and beyond. Organizations like the United Way are an example of how targeted mission statements can have substantial and scalable influence.18

Intermediaries are essential to a comprehensive and coordinated prisoner reentry plan that recruits a large number of skilled and trained volunteers, while developing private and public partnerships to confront key reentry and aftercare problems. Finally, intermediaries are suited to interact with governmental entities while drawing on the substantial human capital of volunteers as well as the social and spiritual capital of individuals and organizations in the private sector. Indeed, the CCF, the Faith and Community Technical Support Project, the Latino Coalition for Faith and Community Initiatives, and Nueva Esperanza are but a few examples of the unique role intermediaries can play at the local and national levels.

29

CONCLUSION

Because prisoner reentry is a problem facing communities across the United States, the ultimate goal of any plan should be to establish a model that is not only effective in a particular area, but one that can be effective on a larger scale in multiple communities. It is one thing to have isolated success, it is quite another to have success at a statewide level. For example, while not a prisoner reentry program, Amachi Texas is a unique public-private partnership designed to reach scale statewide.19 What has been missing until recently is a prisoner reentry model or template that links all the nonnegotiable elements of reentry together in a way that can be replicated and sustained in cost-effective ways in local communities, regionally, or statewide. Thus, a plan is needed where coordination and collaboration are central, where the goals of the reentry model are realistically achievable, where the specific elements of the plan are replicable in a broad range of communities, and finally, where the plan is affordable and does not add new costs to already overburdened correctional budgets.

The role of the government in reentry and aftercare is essential, but it should not be all encompassing. The criminal justice system should be viewed as a key partner among other public and private partners collaborating with the many reentry initiatives being led in the community and coordinated through intermediaries. Ready4Work and the PRI provide preliminary evidence that faith-based and secular organizations as well as national and community government institutions can work together to address comprehensive prisoner reentry in a scalable way. In order to replicate these experiences, the government needs to continue to welcome and accommodate faith-based and community-based volunteers and organizations. Additionally, faith-based and community-based intermediaries will have to bring much-needed expertise in coordinating and training volunteers as well as organizations in the areas of employment, housing, education, and counseling. In this way, faith-based and secular partnerships can play a catalytic role in a comprehensive and scalable approach to prisoner reentry.

NOTES

1. Data are based on face-to-face interviews with 13,986 inmates in 1991 and published by the Bureau of Justice Statistics. Similar surveys were conducted in 1974, 1979, and 1986.

2. On the basis of a multivariate matched sampling method, seven variables most strongly predicted members of the PF groups: age, race, religious denomination, county of residence, military discharge, minimum sentence, and security classification.

3. Founded in 1997, the InnerChange Freedom Initiative is operated by Prison Fellowship Ministries at the Carole Vance Unit outside of Houston, Texas.

30

4. The comparison group was matched with IFI participants based on the following characteristics: race, age, offense type, and salient factor risk score (a correctional assessment tool commonly used in most prisons to help predict the level of risk that prisoners pose to correctional authorities).

5. The subculture of prison inmates has been an ongoing topic of sociological and criminological inquiry. Donald Clemmer coined the term “prisonization” in his book The Prison Community (1958), whereby inmates become socialized into prison culture. An assumption of prisonization is that inmates internalize prison culture, and their subsequent behavior is a reflection of this internalization.

6. Statement by Mark Early, President of Prison Fellowship Ministries, at a White House event on prisoner reentry in March 2007.

7. Baier and Wright (2001) reviewed a total of 60 published studies and found that (1) religious beliefs and behaviors exert a moderate deterrent effect on individuals’ criminal behavior, and (2) conceptual and methodological approaches account for some of the inconsistencies in the research literature. In a second review, Johnson, Tompkins, and Webb (2006) reviewed 151 studies that examined the relationship between religiosity and drug use (n=54) or alcohol use (n=97) and abuse. The majority of these studies demonstrate that participation in religious activities is associated with less of a tendency to use or abuse drugs (87%) or alcohol (94%). These findings hold regardless of the population under study (i.e., children, adolescents, and adult populations) or whether the research was conducted prospectively or retrospectively. In this same study, Johnson reviewed 46 published studies that examined the religiosity-delinquency relationship. Among these studies, 78% reported reductions in delinquency and criminal acts to be associated with higher levels of religious activity and involvements.

8. Executive Order 13199 created the White House Office of Faith-Based and Community Initiatives on January 29, 2001.

9. Executive Order 13198 created five Centers for Faith-Based & Community Initiatives on January 29, 2001; Executive Order 13280 created two Centers for Faith-Based and Community Initiatives on December 12, 2002; Executive Order 13279 requires equal protection for faith-based and community organizations, as of December 12, 2002; Executive Order 13342 created three new Centers for Faith-Based and Community Initiatives at the Departments of Commerce and Veterans Affairs and the Small Business Administration on June 1, 2004; and Executive Order 13397 created a new Center for Faith-Based and Community Initiatives at the Department of Homeland Security on March 7, 2006.

10. Unlevel Playing Field: Barriers to Participation by Faith-Based and Community Organizations in Federal Social Service Programs. Washington, DC: White House Office of Faith Based and Community Initiatives, 2006.

11. The Recidivism Reduction and Second Chance Act authorizes $165 million annually over 2 years to support mentoring programs, substance abuse treatment, literacy classes, job training, and other assistance intended to help ex-offenders pursue productive, crime-free lives after their sentences are up. The bill authorizes grant funding for fiscal years 2009 and 2010 for state and local governments to launch or continue programs to improve ex-

31

offenders' return to society. It also allocates competitive grants to faith-based and community-based nonprofits to offer programs that link ex-offenders with mentors or that help them seek and keep jobs. The bill includes elements of the Bush Administration's Prisoner Re-Entry Initiative (PRI), launched in 2004, which connects ex-offenders with religious and secular nonprofits for mentoring and other programs intended to help them make a successful transition to community life.

12. The 11 sites are: City of Memphis Second Chance Ex-Felon Program (Memphis, TN); Allen Temple Housing & Economic Development Corporation (Oakland, CA); East of the River Clergy, Police & Community Partnership (Washington, DC); Exodus Transitional Community (East Harlem, NY); Holy Cathedral/Word of Hope Ministries (Milwaukee, WI); Operation New Hope (Jacksonville, FL); SAFER Foundation (Chicago, IL); Search for Common Ground (Philadelphia, PA); Union Rescue Mission (Los Angeles, CA); Wheeler Avenue Baptist Church & the InnerChange Freedom Initiative (Houston, TX); America Works Detroit (Detroit, MI).

13. Participant eligibility for Ready4Work was determined based on three factors: (1) age of the ex-offender, (2) presenting offense, and (3) length of time pre- or postrelease. Ex-prisoners between the ages of 18 and 34 who had most recently been incarcerated for a nonviolent felony offense and were no more than 90 days pre- or postrelease were eligible to enroll in the program.

14. For additional examples of federally funded intermediaries, see (1) The Latino Coalition for Faith and Community Initiatives (funded by the Department of Labor), (2) Faith and Community Technical Support (funded by the Office on Violence Against Women), and (3) The Compassion Capital Fund (http://www.acf.hhs.gov/programs/ccf).

15. See Web site for Big Brothers Big Sisters of American (www.bbbsa.org).

16. See Web site for Amachi: People of Faith Mentoring Children of Promise (www.amachimentoring.org).

17. See Web site for American Church Lists (www.americanchurchlists.com).

18. The United Way is a national network of more than 1,300 locally governed organizations that work to create lasting positive changes in communities.

19. Amachi Texas, a joint initiative between the Office of the Governor, the Texas Department of Criminal Justice, Texas Workforce Commission, OneStar Foundation and Big Brothers Big Sisters of Texas, was launched in 2005. The program helps children of prisoners by working in communities throughout the state of Texas via mentoring relationships. The Amachi Texas initiative is currently the subject of a 3-year evaluation that incorporates a randomized controlled study.

32

APPENDIX A

Table A-1. Prisoner Reentry Initiative Sites

Name of Group Location Population Served Organization Type Number of Faith

Partners

1. Primavera Foundation

Tucson, AZ 400 participants Community-Based 1

2. AZ Women’s Education & Employment

Phoenix, AZ 400 male and female ex-offenders released in Maricopa Co, AZ

Community-Based 8

3. Metro United Methodist Urban Ministries

San Diego, CA 200 adult ex-offenders Faith-Based 50+faith-based groups (including

48 churches)

4. Allen Temple Housing & Economic Development Corporation

Oakland, CA 200 adult ex-prisoners Faith-Based 1

5. Mexican American Alcoholism Program Inc.

Sacramento, CA 200 participants Community-Based 1–mentoring

6. Fresno Career Development Center

Fresno, CA 200 clients Community-Based 20

7. Empowerment Program

Denver, CO 200 recently released nonviolent ex-offenders

Community-Based 6

8. Community Partners in Action

Hartford, CT 200 adult ex-offenders; main target men 25–40

Community-Based None known

9. OIC of Broward County

Fort Lauderdale, FL

220 nonviolent ex-offenders

Community-Based 1–(this FBO pulls from 15 different

faith entities), OIC also works with 10 more faith groups

10. Directors Council

Des Moines, IA 210 nonviolent ex-offenders released in the prior 6 months

Community-Based 2–mentoring

11. Safer Foundation

Chicago, IL 300 formerly incarcerated individuals

Community-Based 6

(continued)

33

Table A-1. Prisoner Reentry Initiative Sites (continued)

Name of Group Location Population Served Organization Type Number of Faith

Partners

12. Church United for Community Development

Baton Rouge, LA 200 participants Faith-Based & Community-Based

Many

13. Odyssey House LA

New Orleans, LA 200 adult ex-offenders Community-Based 2–mentoring

14. Span Inc. Boston, MA 200 nonviolent ex-offenders

Community-Based 4

15. Episcopal Community Services of Maryland

Baltimore, MD 300 ex-offenders aged 21–40 with a history of drug/alcohol addiction

Faith-Based 1

16. Oakland Livingston Human Service Agency

Pontiac, MI 200 new parolees Community-Based 2–mentoring

17. St. Patrick Center

St. Louis, MO 200 ex-prisoners over 18 nonviolent, non-sex– related offenses

Faith-Based (Catholic Charities)

7

18. Connections to Success

Kansas City, MO 200 adult ex-prisoners Faith-Based 25-30

19. Career Opportunity Development

Egg Harbor/ Atlantic City, NJ

From Atlantic county Community-Based 1–mentoring

20. The Doe Fund Brooklyn, NY 200 primarily from Queensboro Prison

Community-Based Unsuccessful in partnering

w/FBOs; but utilize many faith-based

mentors

21. Goodwill Industries of Greater NY and Northern NJ

Astoria, Newark, NJ

204 nonviolent offenders Community-Based None

22. Urban Youth Alliance International

Bronx, NY 400 returning prisoners from Rikers Island Jail aged 18–24

Faith-Based 1

23. Talbert House Cincinnati, OH 200 ex-prisoners Community-Based 3

(continued)

34

Table A-1. Prisoner Reentry Initiative Sites (continued)

Name of Group Location Population Served Organization Type Number of Faith

Partners

24. S.E. Works Portland, OR 200 participants Community-Based No mention of faith-based

partners

25. Connection Training Services

Philadelphia, PA 208 recently released ex-offenders each year

Community-Based 2

26. WABC Central City Comprehensive Community Center

Houston, TX Ex-offenders in Harris Co, TX

Faith-Based 4

27. Goodwill Industries of San Antonio

San Antonio, TX 400 returning offenders Community-Based 2

28. Urban League Greater Dallas & North Central TX

Dallas, TX 400 participants Community-Based 10

29. People of Color Against AIDS Network

Seattle, WA 200 nonviolent ex-offenders & subgroup of violent offenders of color

Community-Based 10

30. Word of Hope Ministries

Milwaukee, WI 200 formerly incarcerated individuals

Faith-Based 1

REFERENCES

Baier, C., & Wright, B. (2001). If you love me keep my commandments: A meta-analysis of the effects of religion on crime. Journal of Research in Crime and Delinquency, 38, 3–21.

Bauer, L. (2002). Justice expenditure and employment in the United States. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.

Brooks, A. C. (2006). Who really cares: The surprising truth about compassionate conservatism. New York: Basic Books.

Clain, S., & Zech, C. E. (1999). A household production analysis of religious and charitable activity. American Journal of Economics and Sociology, 58, 923–946.

Clear, T., & Sumter, M. (2002). Prisoner, prison, and religion: Religion and adjustment to prison. Journal of Offender Rehabilitation, 35, 127–159.

35

Clemmer, D. (1958). The prison community. New York: Rinehart & Company.

Cornwell, B., & Harrison, J. A. (2004). Labor unions and voluntary association membership. American Sociological Review, 69, 751–767.

Cullen, F. (2002). Rehabilitation and treatment programs. In J.Q. Wilson, & J. Petersilia (Eds.), Crime: Public policies for crime control (pp. 253–289). Oakland, CA: ICS Press.

Evans, T. D., Cullen, F., Burton, V., Dunaway, R. G., Payne, G., & Kethineni, S. (1996). Religion, Social Bonds, and Delinquency. Deviant Behavior 17, 43–70.

Evans, T. D., Cullen, F. T., Dunaway, R. G., & Burton, V. S., Jr. (1995). Religion and crime reexamined: The impact of religion, secular controls, and social ecology on adult criminality. Criminology, 33, 195–224.

Farley, C., & Hackman, S. (2006). Ready4Work in brief—Interim outcomes. Recidivism at half the national average. In P/PV in Brief. Philadelphia: Public/Private Ventures.

Farley, C., & McClanahan, W. (2007). Ready4Work in brief—Update on outcomes; Reentry may be critical for states, cities. In P/PV in Brief. Philadelphia: Public/Private Ventures .

Fink, B., & Branch, A. (2005). Promising practices for improving the capacity of faith- and community-based organization. Philadelphia: Branch Associates, Inc. & Abt Associates, Inc.

Finn, P. & Kuck, S. (2003). Addressing probation and parole officer stress. Final Report to the National Institute of Justice. Retrieved from http://www.ncjrs.org/pdffiles1nij/ grants/207012.pdf

Freeman, R. B. (1986). Who escapes? The relation of churchgoing and other background factors to the socioeconomic performance of black male youths from inner-city poverty tracts. In R.B. Freeman & H. J. Holzer (Eds.), The black youth employment crisis. Chicago: University of Chicago Press.

Hercik, J. M. (2005). Rediscovering Compassion: An Evaluation of Kairos Horizon Communities in Prison. Caliber Associates.

Honig, M. I. (2004). The new middle management: Intermediary organizations in education policy implementation. Educational Evaluation and Policy Analysis, 26, 65–87.

Independent Sector. (2001). Giving and volunteering in the United States 2001: Findings from a national survey. Washington, DC: Author.

Jang, S. J., & Johnson, B. (2001). Neighborhood disorder, individual religiosity, and adolescent drug use: A test of multilevel hypotheses. Criminology, 39, 501–535.

Jang, S. J., Johnson, B., & Bader, C. (in press). The cumulative effect of religiosity in crime desistance. Journal of Drug Issues.

36

Johnson, B. (2002). Assessing the impact of religious programs and prison industry on recidivism: An exploratory study. The Journal of Corrections, 28, 7–11.

Johnson, B. (2003). The InnerChange freedom initiative: a preliminary evaluation of a faith-based prison program. Waco, TX: Baylor University, Institute for Studies of Religion.

Johnson, B. (2004). Religious program and recidivism among former inmates in prison fellowship programs: A long-term follow-up study. Justice Quarterly, 21, 329–354.

Johnson, B. (2007). A Tale of two religious effects: Evidence for the protective and prosocial impact of organic religion. In K. Kline (Ed.), Authoritative communities: The scientific case for nurturing the whole child. New York: Springer Press.

Johnson, B. (2008). The role of religious institutions in responding to crime and delinquency. In P. B. Clarke (Ed.), The Oxford handbook of the sociology of religion. Oxford, England: University of Oxford Press.

Johnson, B., Jang, S. J., Larson, D., & Li, S. (2001). Does adolescent religious commitment matter? A reexamination of the effects of religiosity on delinquency. Journal of Research in Crime and Delinquency, 38, 22–44.

Johnson, B., Larson, D. B., & Pitts, T. (1997). Religious programming, institutional adjustment and recidivism among former inmates in prison fellowship programs. Justice Quarterly, 14, 145–166.

Johnson, B., Larson, D., Jang, S. J., & Li, S. (2000a). The “invisible institution” and black youth crime: The church as an agency of local social control. Journal of Youth and Adolescence, 29, 479–498.

Johnson, B., Larson, D., Jang, S. J., & Li, S. (2000b). Who escapes the crime of inner-cities: Church attendance and religious salience among disadvantaged youth. Justice Quarterly, 17, 701–715.

Johnson, B. R., Tompkins, R. B., & Webb, D. (2006). Objective hope: assessing the effectiveness of faith-based organizations: A review of the literature. Waco, TX: Baylor University, Institute for Studies of Religion.

Langan, P. A., & Levin, D. J. (2002). Recidivism of prisoners released in 1994. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. NCJ 193427.

McClanahan, W. (2007). P/PV Preview: Mentoring ex-prisoners in the Ready4Work reentry initiative. Philadelphia: Public/Private Ventures.

McGowen, R. (1995). The well-ordered prison: England, 1780–1865. In N. Morris & D. J. Rothman (Eds.), The Oxford history of the prison: The practice of punishment in western society. New York: Oxford University Press.

37

Musick, M., & Wilson, J. (2007). Volunteers: A social profile. Bloomington: Indiana University Press.

Osborne, J., & Solomon, A. (2006). Jail and reentry roundtable initiative. Washington, DC: Urban Institute.

Peters, E.M. (1995). Prison before the prison: The ancient and medieval worlds. In N. Morris and D. J. Rothman (Eds.), The Oxford history of the prison: The practice of punishment in western society. New York: Oxford University Press.

Petersilia, J. (2003). When prisoners come home: Parole and prisoner reentry. New York: Oxford University Press.

Putnam, R. D. (2001). Better together: Report of the Saguaro Seminar. Cambridge, MA: Harvard University, John F. Kennedy School of Government.

Travis, J. (2005). But they all come back: Facing the challenges of prisoner reentry. Washington, DC: Urban Institute Press.

Travis, J., & Visher, C. (Eds.). (2005). Prisoner reentry and crime in America. Cambridge University Press (pp. 1; 255–256).

U.S. Department of Justice. (2006). Adults on probation, in jail or prison, and parole. Sourcebook of Criminal Justice Online. Table 6.1. Retrieved from http://www.albany.edu/sourcebook/pdf/t612006.pdf

38

IMPROVING ORGANIZATIONAL CAPACITY OF FAITH-BASED AND COMMUNITY-BASED

DOMESTIC VIOLENCE SERVICE PROVIDERS

Mitchell Brown

In 2006, members of a small church in rural southwest Idaho decided to go door-to-door to speak with area residents, most of whom were Spanish-speaking immigrants, to reach out and invite the mostly migrant workers to visit their church. Unexpectedly, they encountered numerous women who disclosed in the course of conversation that they had been victimized by domestic violence and were unable to connect with others in the community because of language barriers and a lack of transportation. The church pastor, himself a recent immigrant with poor English skills, reached out to Catholic Charities.

Around the same time, a small domestic violence service and advocacy organization in a nearby county received a grant from the U.S. Department of Justice, Office of Violence Against Women (OVW) to be an intermediary organization in a pilot program designed to make connections between domestic violence services and the faith community in rural areas. As an intermediary, this organization would assist OVW in attracting grant applicants, oversee the work of the funded applicants, and provide technical assistance (TA). Representatives of the newly designated intermediary organization also contacted Catholic Charities to ask if they knew of any churches that were working with victims of domestic violence. The church in southwest Idaho and the intermediary began working together, although this was complicated by the fact that neither the church nor its pastor had e-mail or Internet access. The overall goal of this collaboration was to help the church develop an outreach program for Spanish-speaking migrant women. It took a significant amount of time to build a trusting relationship between the two entities. Once that relationship was established, it took additional time to help the church apply for a small OVW grant to fund these efforts.

Although recommended by the intermediary organization for funding, ultimately the project was not funded by OVW. Nonetheless, the process and project highlight some of the possibilities and challenges of Charitable Choice, the provision included in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, a welfare reform law that clarified that faith-based and community organizations (FBCOs) could compete for federal funding on the same basis as other groups. Charitable Choice came to represent a key element of President Bush’s Faith-based and Community Initiatives (FBCI). (For more information on the history and debate surrounding

39

Charitable Choice and the FBCI, see, for example, Ackerman & Burke, 2001; Black, Koopman, & Ryden, 2004; Carlson-Thies, 2003; Formicola, Segers, & Weber, 2002.)

This paper focuses on the organizational capacity issues that arose during the evaluation of the Rural Domestic Violence and Child Victimization Enforcement Grant Program Special Initiative: Faith-based and Community Organization Pilot Program—also known as the Rural Pilot Program (RPP)—funded by OVW. Four main questions are addressed: (1) What were the capacity needs of the pilot organizations participating in this program? (2) What were the differences between the faith-based and community-based pilot organizations? (3) To what extent does an intermediary model improve the capacity of faith-based versus community-based organizations? and (4) Do different forms of intermediary technical assistance delivery impact organizational capacity differently?

First, the paper presents the background on the RPP initiative, the needs in rural areas that gave rise to this program, and the importance of organizational capacity. This is followed by a detailed description of the evaluation, the data, the analyses, and the findings. Finally, the implications of the findings for policy, research, and practice are discussed.

BACKGROUND

The RPP was designed to fill a gap in the provision of social services in rural areas. Most rural areas suffer from a lack of resources, low levels of economic growth, and limited service provision. Further, victims of domestic violence in rural areas experience an increased incidence of sexism and racism (Websdale, 1998). Service providers that do exist in rural areas often operate on shoestring budgets, with volunteers providing significant amounts of work despite limited availability of training and a reliance on personal rather than organizational resources.

To try to address these needs, the RPP provides small grants ($10,000–$100,000) to faith-based and community-based organizations in rural areas to provide domestic violence services through an intermediary model over a one-year period.1 OVW intended this grant program to enable small grassroots organizations in rural areas to “create new avenues of partnership and collaboration between small faith-based and community organizations and law enforcement officers, prosecutors, victims’ advocacy groups, and other related parties” (OVW, 2005).

OVW selected an intermediary model to implement the RPP. This type of model is one in which a federal agency selects one or more organizations to fund, and these organizations in turn issue their own Request for Proposals (RFPs), receive and review funding applications, oversee subawardee work, provide TA, coordinate reporting, and receive and respond to reimbursement

40

requests. As federal functions are devolved to the state and local levels, intermediary organizations are seen as providing important support to FBCOs. Providing TA under this model to subaward FBCOs has been limited because of geography and funding levels (Hall, 2004; Klein & Wilson, 2005; Miller, 2004).

Three intermediary organizations were selected for the RPP. They were limited to the use of 20% of their overall budget for project administration and TA provision. One intermediary, Advocates Against Family Violence, provided subawards initially only in southwest Idaho; this organization is referred to as the regional intermediary. A second intermediary, the Wyoming Coalition Against Domestic Violence and Sexual Assault, provided subawards throughout Wyoming; this organization is referred to as the state intermediary. A third intermediary, the Faith and Community Technical Support Program, sponsored by Baylor University and the Montana Office for Victims, provided subgrants to rural communities across the nation; this organization is referred to as the national intermediary.2 Each of the intermediaries provided different forms of TA and capacity support to their subgrantees. Only one organization, the state intermediary, required formal partnerships between both faith-based and community-based organizations to qualify for the grant.

Overall, the three intermediaries funded 54 subgrantees out of 176 applicants.3 (For a list of the funded organizations, see Appendix A.) Of the applicants, two thirds were community-based organizations and one third was faith-based organizations. Only 31% of the applicants were funded. These organizations are spread around the country, but all are in rural jurisdictions.

Understanding Capacity Issues of FBCOs

From the inception of this program, OVW expected that many of the FBCO applicants would have limited capacity and would require TA in many areas, from effective domestic violence programming to organization operations to organizational aspirations and strategy (OVW RPP Request for Proposals, 2005). Further, it was anticipated that faith-based organizations in particular would need assistance, a view that stemmed from the heterogeneity of these groups (Chaves, 1993); that is, the fact that religious organizations have varying types and levels of formalized organization that react differently to outside influences (Benson & Dorsett, 1971; De Vita & Wilson, 2001; Harris, 1998) as well as a reluctance among some faith traditions to acknowledge or support victims of domestic abuse (Lasco, 2001) because of the perception that to do so works against the goal of keeping marriages and families together.

Organizational capacity refers to a myriad of areas, including both functional capacity and the capacity to provide services. Typically, capacity comprises the following components:

41

management and operations, board of directors and governance, key allies, resources, program planning and implementation, and evaluation. A key question that drives an interest in organizational capacity is to determine what is the best way to ensure the sustainability of programs and organizations after federal funding ceases.

Six key factors have been found to be important to sustainability across different types of community-based organizations:

# the ability to develop a vision or legacy (Center for Mental Health in Schools, 2001; Cutler, 2002; Hayes & Bryant, 2002; Kaufman, 2002; Wolff, n.d.);

# appropriate planning, evaluation, and commitment of key allies (Cutler, 2002; Hayes & Bryant, 2002; Kaufman, 2002; Kubisch et al., 2002; Schorr, Sylvester, & Dunkle, 1999; Wolff, n.d.);

# the capacity to create and strengthen collaborations and partnerships (Kubisch et al., 2002; Wolff, n.d.);

# sufficient funding and other forms of resources (Hayes & Bryant 2002; Schorr et al., 1999; Staggenborg, 1988);

# the ability to influence policy and change norms (Hayes & Bryant, 2002; Schorr et al., 1999; Staggenborg, 1988); and

# the ability to adapt to changing contextual factors (Hayes & Bryant, 2002; Imig, 1992; Minkoff, 1999; Schorr et al., 1999).

However, many challenges to sustainability have also been identified, including geography (Wolpert & Reiner, 1984); low levels of community social capital (Gomez & Santore, 2001); organizational size, age, and transitions (Minkoff, 1993); and tensions in collaborations around funding and planning (Banaszak-Holl, Allen, Mor, & Schott, 1998).

The evaluation team anticipated that these factors applied equally to faith-based organizations and secular community-based organizations, and that if the RPP intermediaries were successful in their service delivery, at a minimum by the end of the grant period the subawardees should be able to

# obtain funds, beyond the Pilot Program grants, that they most likely would not have otherwise obtained;

# make new connections to sources of volunteers and sources of in-kind support;

# access relevant TA that would not otherwise be obtained;

# improve their social marketing skills in order to increase exposure to grassroots groups, the public media, and endorsements;

42

# improve their organizational, administrative, and fiscal controls with such changes as approval as a 501(c)(3) nonprofit corporation; and

# develop regular intensive and extensive organizational mentoring and coaching relationships with the intermediary (Wilson et al., 2005).

DATA AND METHODS

The evaluation of the RPP includes a process evaluation, outcome evaluation, study of the value-added of a faith component, and a capacity study. The evaluation of capacity-building in this pilot program, which is the focus of this paper, was conducted using multiple approaches (also called a mixed-method design) in order to increase confidence in the findings. Capacity components were examined in six areas: management and operations, board of directors and governance, key allies, resources, program planning and implementation, and evaluation (see Appendix B). These areas comprise the most frequently talked about and used aspects of organizational capacity in capacity studies and were developed in part through an adaptation of the McKinsey and Company (2001) capacity assessment report and tool, as well as through the experiences of the Institute for Community Peace (ICP), a national violence prevention organization begun in 1994 as a public-private partnership, in working with grassroots community-based organizations to prevent an array of forms of violence. The questionnaire, focus group protocols, site visit protocols, and monthly telephone interview protocols discussed below were all developed around each of these areas.

To assess the effectiveness of the capacity-building portion of the RPP, a triangulated approach was used that included case studies of a subset of funded pilot sites (four from the national intermediary, two from the state intermediary, and two from the regional intermediary); precapacity and postcapacity assessments of all applicants, whether funded or not, which were delivered on-line (or on paper for those organizations unable to fill out an on-line survey); and focus groups conducted with all of the funded organizations. In addition, the intermediary organizations were interviewed monthly for 18 months and an OVW representative was interviewed once. All respondents were assured confidentiality.

Case Study Sites

Case study sites were chosen to profile the different issues faced by faith-based as opposed to community-based grantees and newer as opposed to established organizations. Only organizations that received funding were included in the case studies. During the course of the grant, six of the eight case study sites were visited (two of the sites twice), and monthly telephone calls were held with each site.4 The one- to two-day site visits were used to assess the organization’s strengths and needs as well as their capacity to provide domestic violence

43

services. For each organization, the project director and other organization staff were interviewed. The site visits were also used for document collection. The monthly calls covered the activities of the organizations, the different capacity-building supports that the subawardees received, and their successes and challenges.

On-line Assessment

An on-line, self-administered capacity assessment was sent to each of the applicants, whether funded or not, at the start and at the end of the subgrant year. Organizations that could not fill out the on-line survey because of technical limitations were given the option of filling out and returning a paper version or being “interviewed” via telephone with an ICP staff member who then filled out the on-line survey for the respondent. The response rate was 90.2% for preassessment funded organizations, 48.8% for preassessment nonfunded organizations,5 98% for postassessment funded organizations, and 28.8% for postassessment nonfunded organizations. Although the lower response rates from the nonfunded organizations are disappointing, they are not atypical of survey response rates in general. Nonetheless, the possibility exists that this lower response rate introduces some bias into the data, raising the possibility that the findings about the nonfunded organizations are not necessarily representative.

Each organization that filled out the assessment was entered into a raffle for a $35 gift certificate, and five winners were selected for each of the pre- and postassessments. To analyze these data, a combination of descriptive statistics was used as well as measures of change by type of subawardee (funded versus unfunded, intermediary type, and faith-based versus community-based) between the start and end of the study. The analyses reported in the findings section include cross-tabulations, correlations, and difference of means and difference of proportions tests.

Focus Groups and Interviews

The evaluation team convened a national meeting in Denver, CO, in September 2006, at which focus groups were held with the subawardees to discuss their capacity to carry out their work, areas of strengths and needs, the utility of the capacity support provided, and suggestions for program improvements. These data were used to augment findings from the case studies and capacity reports.

In-depth conversations with the intermediaries were held at the start of the grant period, as well as monthly telephone conversations throughout the grant period. The first conversation focused on capacity-building plans, whereas the actual services provided and the intermediaries’ perspectives on what did and not work well were discussed during the monthly conference calls.

44

In addition, each of the subawardee organizations was surveyed about their experiences with the intermediaries, particularly around the areas in which they received support and the value-added of the support received. The response rate for this survey was 88%. Finally, on-site visits to the offices of two of the intermediaries at the start of their grant were used to discuss their programs; for the third intermediary, this process and the information presented were observed at one of their regional TA meetings.

FINDINGS

Among the organizations that responded to the capacity assessment, organizational age ranged from new at the start of the grant to 56 years, with a median starting year of 1997; suggesting that many, if not most, of the organizations were relatively young. There was no statistically significant relationship between organizational age and whether or not an organization was funded, contrary to the general belief that organizational age is a predictor of success in competing for grants.

These organizations were all located in rural jurisdictions around the country, although more were located in southwest Idaho and Wyoming because of the regional and state intermediaries in those areas. A third of the applicant organizations were faith-based. The primary programs and activities of the applicants included victim and legal advocacy, case management, and provision of shelters and counseling services, although there was a wide range of activities beyond these, including outreach to churches and the development of training programs to address faith-based approaches to providing domestic violence services. There was no statistically significant difference in the services provided between the successful and unsuccessful applicants.

Capacity Needs of the Organizations

All of the organizations that applied for funding were in need of TA and support. However, the types of assistance needed differed from that anticipated by the original RFP. OVW expected that many of the subgrantees would need assistance with very basic operational functions, such as obtaining 501(c)(3) status. However, the application process screened out all but one of the 13 organizations that did not have 501(c)(3) status; in other words,12 of the 13 organizations without nonprofit status at the time of the application were unsuccessful in securing funding.6,7

The areas in which the organizations needed the least assistance (as determined by self-report) included (1) identifying benchmarks for evaluation (4.9% reported that they needed help in this area), (2) acquiring in-kind resources (4.7%), (3) developing a vision/mission statement (2.3%), (4) volunteer management (2.3%). and (5) obtaining 501(c)(3) status (2.3%). All of these, except

45

one, corresponded with observations from site visits. All of the organizations visited had a nonprofit status and had previously identified mission and/or vision statements. Also, they were all skilled at managing volunteers and acquiring in-kind resources—a reflection of the limited resources with which they operated in isolated areas, requiring that they be creative in attracting “free” resources.

The only disconnect in these findings that emerged from the case studies concerned identifying benchmarks for evaluation. In large part, the organizations, other than counting the number of people served, were unfamiliar with evaluation strategies8 and techniques, making this an area in which they all needed assistance, but only a handful recognized this need. The intermediaries did no better in identifying or responding to this. The one intermediary that had an in-house evaluation capacity had decided to provide primarily substantive as opposed to organizational assistance, whereas the other two would have needed to hire outside experts to assist in this had they seen it as a priority, which they did not.

Areas of Greatest Need

The areas of greatest need, as reported by the subawardees, are more telling and include the following: (1) reporting to OVW (59.1%), (2) sustainability planning (38.1%), (3) budgeting (32.6%), (4) developing or changing operating policies (31.8%), and (5) staff training and development (27.9%). Universally, the subaward case study sites and the intermediaries conveyed reporting problems to OVW. The program was designed to make funding more accessible to small, resource-poor organizations with no prior experience applying for federal funding. As a continuation of this logic, the subawardees were not required to use the standard OVW reporting procedures; thus, new procedures were developed. The problem with this strategy was that the reporting requirement decisions were not made prior to the award decisions, and when made, the requirements were not uniform across the intermediaries. Related to reporting, approximately one third of the responding subawardees indicated that they needed assistance with budgeting; this was primarily related to questions and concerns about reclassifying costs and changing amounts within budget categories. All of these problems were satisfactorily resolved through consultation with the intermediary organizations.

Over one third of the organizations reported needing support in developing sustainability plans. Each of the intermediary organizations took a different approach to addressing this need. One intermediary used a listserv to distribute grant announcements to the subawardees as the announcements came in, although they did not provide assistance in writing or reviewing grant applications because of conflict of interest concerns. Another intermediary took a similar

46

approach but did assist with reviewing grant applications. The third intermediary collaborated with subawardees in putting together an application for a large federal grant program that was funded. However, after the award was made, decisions about how the monies were to be used and distributed produced disagreements among the intermediary and subawardees that ultimately resulted in arbitration.

Finally, approximately 28% of the subawardees reported needing assistance with staff training and development, which extended to volunteers as well. From observations and conversations with the case study sites and the intermediaries, this need was a consequence of operating in rural areas with limited resources. In rural areas where programs are operated by one or two staff members and a handful of volunteers, an individual taking a day away from work requires finding someone else to cover that person’s position in the office or court, and money for travel and training fees, all of which are scarce. One of the intermediaries responded by providing telephone assistance, and sometimes bringing in expert consultants, to handle issues on a case-by-case basis. Another intermediary sent staff to the subaward sites and provided new staff and volunteer training, saving the organization staff time and resources. However, only a portion of the need that was expressed by the subaward organizations was met.

Differences Between Faith-Based and Community-Based Organizations

Few differences were found between the faith-based and community-based organizations in the assessments, case studies, and focus groups. Where there were differences, they appeared in three areas: governance, funding streams, and volunteerism. Faith-based organizations were less likely than community-based organizations to have a board of directors overseeing their operations.9 Of those organizations that do have such boards, there were some important differences between the groups concerning board membership. The faith-based organizational boards have fewer domestic violence experts,10 but more clergy, ministers, and staff.11 Although not unexpected, these differences have implications for each type of organization. For example, faith-based organizations have less expertise in domestic violence, whereas community-based organizations have fewer connections to the faith community.

With respect to funding streams, the faith-based organizations were less likely to have received government funding prior to applying for this grant.12 While this was true for the faith-based applicants that received funding from this program, the relationship was stronger for organizations whose applications were not approved.13 This finding highlights two key points. First, to the extent that one of the goals of the RPP was to reach faith-based organizations outside of mainstream government funding, the program, at least in this iteration, has been successful.

47

Second, these organizations were nonetheless not as competitive as the community-based organizations, suggesting that more might be done to support novice, particularly faith-based, organizations approaching this type of funding stream.

In regard to volunteerism, while faith-based organizations were, naturally, more able to recruit from churches,14 the community-based organizations were much more likely to recruit through newspapers, radio, and television advertisements.15 This translates into a real difference between the two types of organizations in the number of volunteers they were able to recruit. The median number of volunteers for faith-based organizations at the start of the program was 29; the number for the community-based organizations was 15.16 Thus, despite the costly measures some of the community-based organizations use to recruit volunteers, the faith-based organizations attracted significantly greater numbers of volunteers, possibly because they have an engaged volunteer pool at their immediate disposal.

Improving the Capacity of FBCOs Using the Intermediary Model

OVW required that 80% of the funds awarded to each intermediary organization be passed to the subawardees, such that only 20% could be retained by the intermediary organization for program administration and TA costs. Nonetheless, the intermediary organizations were able to provide a great deal of support to the subawardees, and the majority of subawardees reported that the TA they received met or exceeded their needs. Most of the subawardees perceived the evaluation meeting in Denver to be TA and were satisfied with the content of that TA as well. Because of the rural nature of the work of these organizations, simply connecting with others doing similar work served as assistance and support.

Positive changes in organizational capacity occurred within many of the funded organizations as compared with organizations that were not funded, although these changes were small. This finding was understandable given the short timeframe of the grant period. With 6 to 12 months to implement new or expanded programs, while at the same time working to increase organizational capacity to provide services and bolster organizational strength and sustainability, the subawardees were unlikely to display much difference, despite the good works and best efforts of the intermediary organizations.

Preassessment and Postassessment Changes

Examining preassessment and postassessment changes by whether an organization was funded or not, there were changes in only a few areas. Appendix C displays the results only for those items that were statistically significant and not for those for which there was no change between the pre- and postassessments. Most of the positive changes were in the areas of management and

48

operations, boards of directors, and governance. The more interesting issue, however, is to what extent these changes came from the receipt of the grant itself versus assistance provided by the intermediary organizations.

The organizations that were not funded experienced few changes, but where there were changes, they were primarily negative. These organizations saw decreases in representatives from various sectors on boards of directors, including community members, community leaders, field experts, and clergy and ministers. They recruited volunteers from fewer outlets over the period, and some organizations that had previously relied on a volunteer base lost those as well. At the same time, however, they realized some gains. Nonfunded organizations increased managerial infrastructure during the period by developing human resource policies and management information systems. More importantly, there was an increase during the study period in the number of these organizations that received government funding. It makes sense that there might be such an increase for the organizations that were funded, but it does not necessarily follow that those that did not receive funding would also experience such increases. While it is not possible to know for certain given the current data, it may be that there was something about the process of applying for this grant program that either encouraged applicants, even unsuccessful ones, to pursue other government funding, or helped them to develop important local contacts to facilitate the process. Given the available data, other explanations cannot be ruled out, but this may potentially be a positive and unintended consequence of the program that deserves greater scrutiny in the future with similarly constructed programs.17

There were many more positive changes among the funded organizations over the grant period. First, subawardees increased their information technology capacity, with greater availability of e-mail (for organization employees) and the development of Web sites. Second, staff size increased among these organizations. This was the case for both the faith-based and community-based organizations. Both of these types of increases are directly attributable to the grant itself and not to the TA these organizations received—the increased funding allowed the organizations to pay for e-mail accounts, pay for someone to develop Web sites, and fund staff. To the extent that the case study sites are representative of the other funded organizations, much of the grant funds were used to hire paid staff among the individuals who had previously worked as volunteers. These organizations also realized increases in different segments represented in their boards of directors over the study period, notably community members and field experts. Rather inexplicably, the number of lawyers serving on boards decreased at the same time.18

Despite these capacity increases for the funded organizations, there were also areas in which the funded organizations appeared to experience a decrease in certain capacities. Each of these areas

49

reflected the more sophisticated components of organizational capacity, such as having a theory of change, strategic plan, fundraising plan, or having conducted asset mapping. In discussions with the case study sites, these organizations revealed that they did not fully understand some of the questions in the preassessment survey. Through exposure to materials in these areas given by the intermediaries, the evaluation meeting, and in interactions with the evaluation team, the organizations increased their knowledge (but not practice) and did not actually revert on these items through the course of the grant period. In short, they answered more accurately in the postassessment than in the preassessment survey.

Differential Impact of Intermediary Technical Assistance Delivery on Organizational Capacity

Each of the three intermediaries provided TA in different ways and focused on different issues. The regional model implemented monthly meetings, which grantees were required to attend, that covered a range of substantive topics led by experts and also provided telephone or in-person TA on request, primarily about operational issues. The statewide model hired a full-time TA provider who regularly traveled to the different programs around the state and responded to telephone requests concerning substantive and operational issues. The national model divided the country into regions and held monthly conference calls with each region as well as one regional meeting, and in addition frequently provided resources to the subawardees via a listserv. The materials provided primarily concerned substantive issues.

Each of these delivery models possesses strengths and limitations. The national intermediary brought greater experience, contacts with national experts and funding, and the ability to leverage these resources to craft a TA program that included frequent information exchange via listserv and conference calls as well as the ability to respond to individual problems through their consultant network. The limitation of this model was that the national intermediary had fewer personal contacts with its subawardees, restricted to the meeting in Denver, its regional meetings, and visits to approximately one third of the subaward sites.

The statewide intermediary hosted a training meeting at the start of the award period that was open both to the programs that were funded and those that were not. Following that meeting, all other TA was provided through telephone calls and regular visits to the program sites by a staff member whose time was dedicated to providing assistance. While the subaward programs received more face-to-face time with the state TA provider, the content of that TA was less sophisticated than that provided by the national model.

50

Finally, the regional program, with the regular, mandatory contacts appeared to be the most promising of the models based on the amount and quality of contact and feedback from the subawardees. This system, however, faced significant challenges toward the end of the grant period because of a conflict between the subaward organizations and the intermediary. The intermediary organization partnered with the subaward organizations to cooperatively apply for federal funding through a different grant program and succeeded in obtaining that funding. However, when the new project was implemented, the subaward organizations felt they had been verbally promised resources that the intermediary organization did not deliver. The subaward organizations later lost this challenge in arbitration, but the conflict was avoidable. While the responsibility for the conflict resided with the subaward organizations for failing to read the fine print before signing on to the grant, the power imbalance between the organizations is such that forbidding intermediaries and their subawardees to enter into such arrangements during the course of a project such as this might be advisable, although such a rule could have far-reaching implications for many federal programs, and might not be welcomed by many subawardees.

A key question is whether one or another of these intermediary approaches translated into greater improvements for subawardees. Based on the current findings, the answer is not promising. None of the pre-post assessment differences could be tracked back to the different intermediary models. The only differences among the intermediaries came from the subaward assessments of the utility of the support they received from the intermediaries. Among over 30 capacity areas they were asked to rate, differences among the intermediaries came up in four areas: budgeting,19

technology,20 volunteer recruitment,21 and data collection and management.22 For each of these areas, the national intermediary received higher scores for helpfulness than the state or regional intermediaries.

IMPLICATIONS FOR POLICY, RESEARCH, AND PRACTICE

The evaluation of the RPP adds to a growing body of knowledge about the effects of regulatory reform and the FBCI, and the findings point to several ways to approach the RPP program in the future to enhance its reach and effectiveness. Despite some flaws, this demonstration has increased the capacity of many organizations in rural areas around the country to assist rural victims of domestic violence. The findings presented here also suggest that this pilot may have even helped those organizations that applied but were not successful in securing funding, perhaps by increasing their interest in applying for grant support.

51

FBCI is an example of the positive effects of policy innovation and diffusion. While it is difficult to make specific policy recommendations about working with faith-based and community-based organizations—as not all such organizations are the same nor do they have the same needs—some broad recommendations can be developed from this study. First, to the extent that improving organizational capacity and sustainability is a priority, federal granting agencies should consider committing to longer term funding if they intend for the programs they support to realize improvements in these areas. Second, agencies could improve the intermediary model by providing greater training and resources to intermediaries before the start of the programs, and setting guidelines to avoid conflicts of interest seen in the RPP.

These findings also suggest some lessons for program partnerships in leveraging limited resources. Domestic violence programs have heretofore been organized primarily by community-based practitioners who have historically not worked with the faith community and vice versa. But the RPP has demonstrated that these two sectors are not necessarily incompatible, and when they find common ground, they can extend the reach of their resources, particularly in resource-poor rural communities, and can provide more services to more women given increases in their capacity to do so.

ACKNOWLEDGMENTS

I wish to express my gratitude to the following people for their helpful reviews, suggestions, and assistance: Andy Klein, Mark Small, Rob Fischer, Chris Maloney, Lauren Feldman, Chris Boyles, and Leanne Marek. This project was supported by Grant No. 2005-IJ-CX-0050 awarded by the National Institute of Justice, U.S. Department of Justice to Advocates for Human Potential (AHP). The Institute for Community Peace has been a contractor to AHP on this project.

NOTES

1. The period of the grant program spanned 2005–2007, but the grants themselves were only for one year. One of the three intermediaries experienced a significant delay in their funding approval and so their subawardees implemented their activities at approximately the same time as the subawardees of the intermediaries were approaching the end of their grant periods.

2. The only exception to this was for those areas covered by the Idaho and Wyoming intermediaries.

3. The discussion that follows pertains only to those 82 organizations that responded to both the precapacity and postcapacity assessments.

4. Other case studies were conducted for the outcome portion of the study but are not included here.

52

5. At the time the pretests were administered, applicants did not yet know whether or not their grant applications had been successful.

6. From chi-square test, p<.05

7. It is not known if applicants were screened out intentionally due to their lack of 501(c)(3) status or on some other basis.

8. However, there was no unified approach to counting people that used services. To illustrate, Organization A counted an individual woman and all the services that person used as a “1” on their monthly activity report, while Organization B counted each specific service and interaction they had per client, resulting in a “12” on their monthly activity report. Despite that, the actual number of interactions were exactly the same between the two organizations.

9. From chi-square test, p<.10

10. From chi-square test, p<.10

11. From chi-square test, p<.01 for both

12. From chi-square test, p<.000

13. From chi-square test, p<.05 and p<.01, respectively

14. From chi-square test, p<.05

15. From chi-square test, p<.05, p<.05, p<.10, respectively

16. Difference of means significant at p<.10 with means of 76.2 and 41.3, respectively. Controlling for organization size with staff as a proxy using regression analysis, the coefficient for the faith-based organizations is positive and significant at p<.10.

17. While it is possible that public funding was just generally more available during this period and all organizations reaped the benefits, looking at changes of available funds from OVW during this time, there are only negligible increases.

18. This change was seen primarily in the community-based groups, and there was no statistically significant change in lawyers on boards of directors among the faith-based groups.

19. From chi-square test, p<.05

20. From chi-square test, p<.10

21. From chi-square test, p<.05

22. From chi-square test, p<.05

23. The evaluation team made several requests to OVW for an interview about several issues, including communication with the intermediary organizations. While these requests were not granted, there was a 30-minute conversation with an OVW representative at the Denver meeting. This issue had not yet been brought to light and therefore was not discussed at that meeting.

53

APPENDIX A: ORGANIZATIONS RECEIVING RURAL PILOT PROGRAM SUBAWARDS

Advocacy & Resource Center Sheridan, WY Amherst County Against Domestic Violence Amherst, VA Archuleta County Victim Assistance Program, Inc. Pagosa Springs, CO Blount County CASA Oneonta, AL Boundary County Youth Crisis & Domestic Violence Hotline Bonners Ferry, ID Break the Chain Against Domestic Violence (BTC) Gun Barrel City, TX Carbon County COVE Rawlins, WY CASA Womens Shelter Pine Bluff, AR City Life, Inc Emmett, ID Comite Civico Del Valle, Inc. Brawley, CA Crisis Intervention Services, Inc. Cody, WY Deaf Smith County Crisis Center Hereford, TX Desert Rose Foundation, Inc. Martinsville, IN Douglas Ecumenical Ministries, Inc. Douglas, WY Elijah Haven Crisis Intervention Center, Inc. LaGrange, IN Esther's Haven House, Inc. New Orleans, LA Evergreen Heights Mennonite Church-Advocacy Caldwell, ID First Christian Church–Cheyenne Cheyenne, WY Goshen Co. Task Force on Family Violence & Sexual Assault Torrington, WY Hawaii Family Law Clinic dba Ala Kuola Mililani, HI Hill Country Community Needs Council Fredericksburg, TX Holy Family Catholic Church Booneville, KY Interfaith Hospitality Network of Greater Johnson City Johnson City, TN Iva’s Place, Inc. Lenoir City, TN Jennings County Council on Domestic Violence Vernon, IN Let's Talk, Inc. Chesapeake, VA Macoupin Community Health Centers, Inc. Gillespie, IL Mary’s House (Ministry Alliance for Regaining Your Safety) Pickens, SC Mercy Ministries Shelter for Battered Women and Children Cheraw, SC Monroe County Community Health Access Committee Madisonville, TN Newton County D.A.W.N. Jonesboro, AR North Kingstown United Methodist Church North Kingstown, RI Options, Inc. Monticello, AR Ozer Ministries East Palestine, OH Park County Crisis Center, Inc. Bailey, CO Rock of Ages Caldwell, ID Rose Advocates, Inc. Weiser, ID Rural Enrichment and Counseling Headquarters, Inc. (REACH) Hawley, MN Safe Haven of Person County, Inc. Roxboro, NC Safe Passage, Inc. Melbourne, AR Safe Place Ministries Boise, ID Safeline, Inc. Chelsea, VT San Luis Valley Immigrant Resource Center Alamosa, CO Southwest Arkansas Domestic Violence Center De Queen, AR United Way of Colleton County Walterboro, SC Western Racine County Family Violence Community Coalition Burlington, WI Women’s Crisis Center of Northeast Arkansas, Inc. Jonesboro, AR

54

Women’s Mountain Passages Quincy, CA Women’s Shelter of Hope, Inc. Hialeah, FL Wyoming Association of Churches Cody, WY YWCA of Oak Ridge Oak Ridge, TN Zoe Institute Tahlequah, OK

APPENDIX B:

CAPACITY EVALUATION COMPONENTS OF THE RURAL PILOT PROGRAM

Management and Operations # Accounting policies # Accounting structures # Human resource policies # Operating policies # Operating oversight # Management information system # Non-profit status # Reporting # Work within budget # Managing transitions # Physical infrastructure # Technology # Staff recruitment # Staff training/development # Plan for sustainability

Board of Directors and Governance # Composition # Leadership # Clear role # Board involvement/support # Accountability

Key Allies # Volunteer recruitment # Volunteer training # Volunteer management # Institutional partners # Community partners # Governmental partners # Faith partners (within) # Faith partners (outside)

55

Resources

# Diversified base # Fundraising plan # Grant writing # Grants management # In-kind resources

Program Planning

# Mission/vision # Strategic plan # Program match/goals # Needs assessment # Asset map # Long-term focus # Reflection # Practice theory development # Community education # Prevention # Responsive/effective services # Evidence-based practice # Substantive expertise # Training/qualifications

Evaluation # Theory of change # Program reflects TOC # Outcomes identified # Benchmarks identified # Data collection and management # Mid-course corrections # Process evaluation # Output evaluation # Outcome evaluation

56

APPENDIX C: SUMMARY OF PREASSESSMENT AND POSTASSESSMENT CHANGES

OF THE RURAL PILOT PROGRAM (DIFFERENCE OF PROPORTIONS AND MEANS TESTS)

Component Area Funded Not Funded Management and Operations

# HR Policies + (p<.05)

# MIS + (p<.05) # E-mail + (p<.05) # Web site + (p<.05) # Staff size + (p<.01) Board/ Governance

# Require board financial support

+ (p<.05)

# Rep: Community member + (p<.10) − (p<.10) # Rep: Community leader − (p<.05) # Rep: Field experts + (p<.05) − (p<.10) # Rep: Lawyers − (p<.05) # Rep: Clergy, ministers − (p<.05) Key Allies # No volunteers + (p<.05) # Recruitment methods # Word of mouth − (p<.05) # Church/faith − (p<.01) # News − (p<.05) # Radio − (p<.01) # Internet − (p<.01) − (p<.10) Resources # Fundraising plan − (p<.05) − (p<.05) # Government funding +(p<.001) + (p<.10) Program Planning

# Strategic Plan − (p<.05)

# Asset map − (p<.05) Evaluation # Theory of change − (p<.05)

REFERENCES

Ackerman, D. & V. Burke. (2001). Charitable choice: Background and issues. Huntington, NY: Novinka Books.

Banaszak-Holl, J., Allen, S., Mor, V., & Schott, T. (1998). Organizational characteristics associated with agency position in community care networks. Journal of Health and Social Behavior, 39, 368–385.

57

Benson, K. J., & Dorsey, J. H. (1971). Toward a theory of religious organizations. Journal for the Scientific Study of Religion, 10, 138–151.

Black, A., Koopman, D., & D. Ryden. 2004. Of little faith: The politics of George W. Bush’s Faith-Based Initiative. Washington, DC: Georgetown University Press.

Carlson-Thies, S. 2003. Charitable Choice: Bringing religion back into American welfare. In H. Heclo and W. McClay (Eds.) Religion returns to the public square: Faith and policy in America (pp. 269–298).Baltimore, MD: The Johns Hopkins University Press.

Center for Mental Health in Schools. (2001). New initiatives: Considerations related to planning, implementing, sustaining, and going-to-scale. University of California, Los Angeles: Los Angeles, CA.

Chaves, M. (1993). Intraorganizational power and internal secularization in protestant denominations. The American Journal of Sociology, 99, 1–48.

Cutler, I. (2002). End games: The challenge of sustainability. Baltimore, MD: Annie E. Casey Foundation.

De Vita, C., & Wilson, S. (2001). Faith-based initiatives: Sacred deeds and secular dollars (Brief #5 from the Seminar Series Emerging Issues in Philanthropy). Washington, DC: Urban Institute.

Formicola, J., Segers, M., & Weber, P. (2003). Faith-based initiatives and the Bush Administration: The good, the bad, and the ugly. Lanham, MD: Rowman & Littlefield.

Gomez, R. & Santore, E. (2001). Membership has its privileges: The effect of social capital and neighbourhood characteristics on the earnings of microfinance borrowers. The Canadian Journal of Economics/Revue Canadienne d’Economique, 34, 943–966.

Hall, J. S. (2004). Faith, hope and Charitable Choice in Arizona: 2003. Albany, NY: Rockefeller Institute of Government.

Harris, M. (1998). A special case of voluntary associations: Towards a theory of congregational organization. The British Journal of Sociology, 49, 602–618.

Hayes, C.D., & Bryant, E. (2002). Sustaining comprehensive community initiatives: Key elements for success. New York: The Finance Project.

Healy, K. (2004). Altruism as an organizational problem: The case of organ procurement American Sociological Review, 69, 387–404.

Imig, D. R. (1992). Resource mobilization and survival tactics of poverty advocacy groups. The Western Political Quarterly, 45, 501–520.

58

Kaufman , M. 2002. Building sustainability in demonstration projects for children, youth and families. Washington, DC: Institute for Educational Leadership. Retrieved from http://www.ncjrs.gov/App/Publications/abstract.aspx?ID=193889

Klein, A., & Wilson, D. (2005). Domestic violence shelter and advocacy services. Prepared by Advocates for Human Potential/BOTEC Analysis Corporation. Providence: Rhode Island Justice Commission.

Kubisch, A.C., Auspos, P., Brown, P., Chaskin, R., Fulbright-Anderson, K., & Hamilton, R. (2002). Voices from the field II: Reflections on comprehensive community change. Washington, DC: The Aspen Institute.

Lasco, C. (2001.) Faith-based responses to domestic violence. Washington, DC: National Center for Victims of Crime.

McKinsey & Company. (2001). Effective capacity building in nonprofit organizations. Washington, DC: Venture Philanthropy Partners.

Miller, P. (2004). The public sector in Montana. In Proceedings from the Rural Sociological Society Meeting 2004: Rural challenges to the Faith-Based Initiatives. Columbia, MO: Rural Sociological Society.

Minkoff, D. C. (1993). The organization of survival: Women’s and racial-ethnic voluntarism and activist organizations, 1955–1985. Social Forces, 71, 887–908.

Minkoff, D. C. (1999). Bending with the wind: Strategic change and adaptation by women’s and racial minority organizations. American Journal of Sociology, 104,1666–1703.

Office of Violence Against Women (OVW). (2005). RFP for the rural domestic violence and child victimization enforcement grant program special initiative. Faith-based and community organization pilot program. Application # 2005-93556-MA-IJ. National Institute of Justice.

Schorr, L., Sylvester, K., & Dunkle, M. (1999). Strategies to achieve a common purpose: Tools for turning good ideas into good policies. Washington, DC: Institute for Educational Leadership.

Staggenborg, S. (1988). The consequences of professionalization and formalization in the pro-choice movement. American Sociological Review, 63, 585–605.

Websdale, N. (1998). Rural woman battering and the justice system: An ethnography. Thousand Oaks, CA: Sage Publications.

Wilson, D., Small, M., & Brown, M. (2005). Evaluation of the rural domestic violence and child victimization enforcement grant program special initiative: Faith-based and community organization pilot program. Application # 2005-93556-MA-IJ. (10) National Institute of Justice.

59

Wolff, T. (n.d.). Coalition Building Tip Sheets. Amherst, MA: AHEC/Community Partners. Retrieved from http://www.tomwolff.com/healthy-communities-tools-and-resources.html#pubs

Wolpert, J., & Reiner, T. (1984). Service provision by the not-for-profit sector: A comparative Study. Economic Geography, 60, 28–37.

60

PARTNERSHIPS THAT WORK: PRELIMINARY FINDINGS FROM THE TANF

FAITH-BASED COMMUNITY ORGANIZATION INITIATIVE

Courtney J. Barthle Rebecca M. Makar Jeanette M. Hercik

Traditionally, the services offered by faith-based and community-based organizations (FBCOs) have helped to bridge the gaps that may leave families facing food insecurity, housing instability, and difficult choices between competing needs. Recognizing that multiple issues often challenge low-income working families and individuals, FBCOs have expanded their services to include education and training, child care, transportation, case management, mentoring programs, life-skills training, and victims’ services (Archambault, Kakuska, & Munford, 2001). While many of these small, nonprofit organizations have provided indispensable services to their communities for decades, it at times has been challenging for them to access public funds to support their efforts, primarily because of insufficient capacity to identify funding opportunities or develop competitive grant applications (Hercik & Lewis, 2003).

The potential of FBCOs to play a more significant role in collaborative service provision was bolstered by Section 104 of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), commonly known as Charitable Choice, which disallowed states to favor or discriminate against FBCOs when applying for federal funds to provide social services. When partnering with FBCOs for the provision of social services, states must provide secular alternatives if an individual objects to the religious character of a program. This “alternative service provision” can also be satisfied by providing clients of the Temporary Assistance for Needy Families (TANF) program with service vouchers redeemable at the provider of their choosing.

The number of contracts between states and FBCOs has increased significantly since the implementation of Charitable Choice, and it is generally believed that FBCOs contribute positively to the provision of social services (Sherman, 2000). Often, FBCOs can have a particular impact on hard-to-serve populations facing multiple barriers to employment. For example, FBCOs have made significant strides in providing services to attach ex-offenders to employment (Banks, Hercik, & Lewis, 2004). Carlson-Thies (2000) asserted that Charitable Choice “represents an opportunity for officials who are searching for effective assistance programs…[and] makes it possible for officials to select the most effective programs, whether

61

they are offered by faith-based or secular organizations” (p. 4). In a comprehensive study of nearly 500 welfare-to-work programs in Chicago, Dallas, Los Angeles, and Philadelphia, researchers found that among the 120 faith-based welfare-to-work programs, 50% were receiving government funding (Monsma, 2002).

More recently, studies have shown that when faith-based organizations partner with government agencies, families have greater success in achieving self-sufficiency when compared with other types of service providers. The intersection of government, nonprofit organizations, and faith-based groups can provide the necessary social capital to support individuals on the path from welfare-to-work (Schneider, 2006). Other researchers have reached different conclusions on the issue of faith-based and community-based providers. For example, Kennedy and Bielefeld (2003) found that networks with faith influence were weaker than their secular counterparts and questioned the effectiveness of faith-based job readiness services.

Despite the somewhat mixed findings described above, studies indicate that faith-based programs are generally open to partnerships and possess a high level of credibility within their local communities (Ericson, 2001). Research has stressed the benefits of Charitable Choice, suggesting that many congregations have expanded their services from “commodity-based benevolence” (e.g., food pantries, used clothing centers) to “relational” ministries (e.g., working face-to-face and intensely with needy families) (Sherman, 2000).

One of the aims of the present study is to isolate why faith-based organizations appear to be successful with hard-to-serve populations. Current hypotheses include credibility in the community, access outside of traditional settings and hours, and passionate commitment of volunteers. However, these assumptions have yet to be tested.

Some research findings are particularly relevant to considerations of TANF-funded FBCO services. First, most congregations do not maintain a separate 501(c)(3) to provide social services (Green, 2007), indicating instead that they are providing services directly. Second, FBCOs continue to fill an important role in serving the most disadvantaged persons, even when compared with their secular counterparts (Reingold, Pirogl, & Brady, 2007). The latter point is critically important to the question of the role of FBCOs in serving TANF clients who present with a variety of barriers to successful employment. These clients are the very ones TANF caseworkers often need the most help in serving. And, given the renewed emphasis on work placement contained in the Deficit Reduction Act—the law which recently reauthorized TANF—the capacity of FBCOs to help attach TANF recipients to employment becomes increasingly important.

62

This paper presents the preliminary findings from an ongoing Department of Health and Human Services (HHS) sponsored study of successful TANF-FBCO partnerships. First, overviews of five partnerships are presented, then common themes among them are identified, and finally the implications are discussed for possible future collaborations in social services provision for low-income and at-risk families.

RESEARCH AIMS

The TANF Faith-based and Community Organization Initiative is a multiyear project sponsored by the Department of Health and Human Services, Administration for Children and Families, Office of Family Assistance to study the relationships between TANF agencies and faith-based and community-based social service providers. The inititaive is designed to identify promising practices and disseminate information to TANF and other social service agencies so that they may replicate these types of collaborations. This study includes four discrete research tasks (details are provided in Appendix A).

This paper describes the preliminary findings from the first part of the first task of this effort: the development of a Research Compendium. This task aims to understand the state of the field in terms of TANF-FBCO collaboration, document promising practices for FBCOs and intermediary/cooperative entities, and inform the selection of promising practices for inclusion in the study. The overall study will be completed by early 2010 and the findings will be available from the Welfare Peer TA Network (www.peerta.acf.hhs.gov). The methodology and preliminary findings from the development of the Research Compendium are detailed below.

DEVELOPING THE RESEARCH COMPENDIUM

The Compendium comprises three research elements plus a reporting element and it is designed to narrow the universe of FBCO employment/self-sufficiency service providers from an initial scan of 250 down to approximately 25 selected for in-depth study. Extant research, including that conducted by this research team, will inform this process.1 The three research elements include:

1. An Environmental Scan/Literature Review that focuses broadly on employment-related services provided by FBCOs and will result in a database of 250 potential study sites. This element is the focus of this paper.

2. A Promising Practices review during which criteria for inclusion in the larger study and variables of importance (e.g., urban versus rural sites) will be identified. This step will verify data collected in the literature review and will narrow the scope of potential programs to be studied down to about 40 FBCOs.

63

3. A Readiness Assessment by which the capacity of each potential site to participate in the model-building2 phase of the research study will be examined. It is anticipated that ultimately 25 FBCOs will emerge for continuing research.

Environmental Scan Methodology

To gain a representative profile of TANF-FBCO partnerships, it was necessary to adopt a broad approach to identifying promising practices. The ongoing Environmental Scan draws from a thorough literature review and nominations from the field. The scan was not limited to programs that specifically targeted TANF clients, but rather examined programs with experience and characteristics that could enhance current services for TANF clients or provide a foundation for further innovation. Programs meeting these criteria were identified through consultation with experts in such fields as transportation services, housing programs, and workforce development. The process incorporated a review of published works and discussions with selected sites.

Environmental Scan Findings: Preliminary Profiles

Given the ultimate objective of model and toolkit creation, it is important that the scan include projects representing a wide range of features, such as location, target population, service delivery strategies, services offered, and partnership type and length. Specifically, a range of urban, mid-sized city, and rural models was sought. Moreover, the selected programs offer unique services ranging from family mentoring, faith-based transitional housing, and workforce development, as well as in-kind donations through a network of FBCOs, and an intermediary model working with FBCOs. While more challenging to measure, estimated replicability of practices was another consideration for inclusion. While a diverse set of selection criteria was used, this approach has limitations in that it is a nonrandom sample; thus, generalizability may be limited. Table 1 provides an overview each program included in the study.

Table 1. Overview of Faith-based and Community-based Organizations

Program Location Target

Population

Service Delivery Strategy Services

Partnership Type

Date Established

Faith Connections (Wilson, NC)

Rural TANF Work Mandated

Indirect: requests from DSS

Financial assistance, counseling, auto repair

Contract with TANF

July 2002

Rachel’s House (Columbus, OH)

Urban Female ex-offenders, including TANF Work Mandated

Direct Housing and life-skills support for women reentering society

Contract with TANF

July 2006

(continued)

64

Table 1. Overview of Faith-based and Community-based Organizations (continued)

Program Location Target

Population

Service Delivery Strategy Services

Partnership Type

Date Established

Faith Partners (Colorado Springs, CO)

Mid-size city

TANF Work Mandated

Direct Mentoring for TANF families, intake class for DSS

Contract with TANF

Mentoring and colocation of offices June 1999; class May 2007

Pathways (Petersburg, VA)

Small city

Low-income youths, including TANF Work Mandated

Direct YouthBuild, Workforce Investment Board, and AmeriCorps programming

Contract with WIB and informal partnership with TANF

2001

Seedco (New York, NY)

Urban Organizations serving TANF Work Mandated clients

Intermediary Funding, technical assistance, infrastructure development

Contract with TANF and WIA

2000

Note: TANF=Temporary Assistance for Needy Families; WIB=Workforce Investment Board; WIA=Workforce Investment Act.

Site One: Faith Connections, Wilson, North Carolina

Serving a rural community since July 2002, Faith Connections is a well-established collaborative of 24 interdenominational faith partners and the Wilson County Department of Social Services (DSS). The overall goal of Faith Connections is to assist social workers by providing the additional resources (monetary and nonmonetary assistance) needed to help families and individuals maintain employment and alleviate crisis. Faith Connections’ mission is to provide real-time access to faith community leadership and resources in order to coordinate the assistance given to families in poverty, crisis, or need, serving both TANF clients and non-TANF clients. The offering of in-kind donations and support, and its rural location, makes this program unique among the sites profiled here.

Faith Connections serves as a network of rural churches that acts as a referral agency network to support local DSS clients who are in crisis or need. The Program Coordinator interfaces with Wilson County DSS and the individual faith partners to serve persons in need.

Partnership with TANF Agency and Services to TANF Clients

An interdenominational, faith-based initiative that is housed at Career Plus, a Work First Self-Sufficiency Center, Faith Connections was designed and staffed by the Wilson County DSS to supplement the needs of TANF families. Rather than working directly with families and

65

individuals, Faith Connections’ clients are the referring agencies and social workers themselves. Faith Connections is the liaison between the workers requesting assistance for their clients and the churches who have agreed to assist financially.

Referrals to Faith Connections are not limited to financial assistance. Support can also be provided for spiritual and emotional counseling, pro bono dental work, housing resources, holiday items, food, and other areas that may be identified as a need through the referring agent. Recycled computers (donated by local organizations), furniture, clothing items, cars, and auto repair are other examples of nonfinancial support.

Results and Lessons Learned

Faith Connections leverages state and local government funds to provide families and individuals with comprehensive services and resources to alleviate crisis situations.3 It allows social workers the flexibility to offer more comprehensive solutions for their clients so that anyone in crisis can potentially overcome obstacles and remain on the path to long-term self-sufficiency.

Having served 411 families and individuals in fiscal year 2006–2007, Faith Connections has been able to make a difference in the lives of clients in Wilson. Since July 2002, $88,325 has been received from all the Faith Community partners to help those in need. DSS has made 1,613 total referrals to Faith Connections since July 2002.

The Wilson DSS recognizes the need to allow churches and faith-based partners to maintain their autonomy in determining what services to offer. By maintaining fiscal responsibility at DSS, the Faith Connections program coordinator can focus on the program side and collaborations with the faith-based community. Furthermore, an efficient referral system that allows for maximum accountability has been vital. Faith Connections works with DSS to track each client’s application form to show what services, resources, and programs the client has received. In addition, Faith Connections submits a monthly report to the faith partners summarizing the services, resources, and programs provided to clients with their donations. By locating the Faith Connections program at an off-site building (the Career Plus Self-Sufficiency Center), and not at DSS, the program has become more accessible to clients, services, and faith partners.

Site Two: Rachel’s House, Columbus, Ohio

Rachel’s House is one of numerous programs operated by Lower Lights Ministries, a faith-based not-for-profit organization with extensive experience and success in serving the distressed community of Franklinton, an urban neighborhood on Columbus’ Westside. Rachel’s House was created to support women recently released from prison by providing transitional housing and

66

supportive services. As part of a holistic approach, Rachel’s House focuses on life- skills development, employment attainment and retention, financial management, addiction recovery, and spiritual formation. Rachel’s House works with the local TANF office and provides several mentors to assist each woman in moving toward self-sufficiency.

While Rachel’s House is careful to use the TANF funding appropriately, this faith-based organization does offer specific faith components in their work with clients. Rachel’s House is supported by the presence and support of their sponsoring church, Bellows Church of the Nazarene, and church volunteers play a major role in the success of the program. Rachel’s House also partners with 13 other FBCOs, 6 of which are other churches.

Partnership with TANF Agency and Services to TANF Clients

Rachel’s House receives TANF funding from the Ohio Governor’s Office of Faith-Based and Community Initiatives as well as program support from the Ohio Department of Rehabilitation and Corrections. Since July 2006, Rachel’s House has been receiving TANF funding as part of the Strengthening Ohio’s Families Initiative. Rachel’s House uses the TANF funding only for the needs specified and government funding does not pay for any activities with faith content.4 Rachel’s House also helps residents apply for of food stamps.

The Rachel’s House program serves female ex-offenders from prison populations at Franklin Pre-Release Center in Columbus, and the Ohio Reformatory for Women in Marysville.5 The program begins in prison with a discipleship Bible study. Each week approximately 45 women meet with Rachel’s House volunteers in a small-group setting to discuss life-skills applications. Participants can then apply to be part of Rachel’s House residency on their release.

The Rachel’s House residency program is a continuous program that provides safe, drug-free housing for each participant for 6 to 18 months depending on the individual needs of each participant. While it provides such basic needs as housing and physical wellness, the program is defined by the provision of five specific components: life-skills development, employment attainment and retention, spiritual formation, financial management, and addiction recovery.

After an initial 40-day orientation to Rachel’s House, residents reenter the community with crucial help from staff and volunteers. Working with the women individually to determine their goals, an employment mentor helps each woman decide what kind of job or educational opportunities are most appropriate. A volunteer helps each woman write her resume and focus on job retention. In addition to her employment mentor, each woman is supported by the following:

67

# Personal Mentor: The main purpose of the personal mentor is to be a mature friend who demonstrates a caring interest in the resident.

# Financial Mentor: The financial mentor works with a resident to help her understand budgeting skills, plan for the future, and compare spending habits with targeted objectives.

# Community Support Team: Comprising three to four adults, this team meets monthly with the resident to discuss personal progress, obstacles, goals, and other matters of concern.

Results and Lessons Learned

The programming and commitment at Rachel’s House has resulted in a 10.1% recidivism rate for the 59 women who have been a part of the residency program in its 5-year existence. In a recent large study, the Department of Justice found a three-year recidivism rate of 60% nationally (Deschenes, Owen, & Crow, 2007). Because of their success with Ohio inmates, Rachel’s House has become a model for faith-based transitional homes. To date, five organizations have used the Rachel’s House program model to start similar housing programs for ex-offenders.6

Rachel’s House was able to add more women to their programs as their capacity grew. Location also was a key to success: the residential houses are in the same neighborhood as their sponsoring church, so Rachel’s House had credibility and visibility in the neighborhood. By training and equipping volunteers to serve in specialized roles, the mentors are more effective. Thus, Rachel’s House ensures that the volunteers are using their time and skills in the areas they wish to serve, which is a key element of volunteer retention.

Site Three: Faith Partners, Colorado Springs, Colorado

As a nonprofit organization formed by several faith organizations, Faith Partners provides services to families transitioning from TANF-funded services, and currently collaborates with over 160 churches to do this work. Specifically, by offering volunteer mentoring as well as a 30-day Crossroads class, Faith Partners assists individuals and families leaving welfare who need additional support to make this transition successful and permanent. Working with the El Paso County (Colorado Springs) Department of Human Services (DHS), volunteer mentoring teams are trained to lend encouragement, support, and life-skills coaching to families desiring to become self-sufficient.

Faith Partners has successfully blended a faith-based presence with a skills-based approach. In partnership with churches, they recruit volunteers from the faith community, educate them about poverty and the challenges facing TANF clients, and focus on skill building. Everything Faith Partner volunteers teach and do for the clients is entirely skill based and secular. While

68

volunteers are recruited from communities of faith and the volunteers are motivated by their faith to mentor, they do not actively speak about or encourage the participants to learn about their faith.

Partnership with TANF Agency and Services to TANF Clients

Since June 1999, Faith Partners has contracted with DHS for services to TANF clients, with funding directly from TANF dollars, to connect families and individuals leaving welfare with volunteer mentor teams who can help support that transition. These teams—for whom this work is a 12-month commitment to a specific family—fill a gap in the lives of these families and help prevent welfare recidivism by providing soft skills training and a solid foundation of support during the transition period. In addition to being trained in the issue areas listed below, mentoring team members are given resources for sharing with mentees. DHS relies on referrals and the widespread distribution of materials to clients about Faith Partners services to accomplish this mentoring goal, as well as referring clients directly to Faith Partners. Faith Partners questions the family about their needs in the mentoring relationship; this intake process drives the creation of a mentoring team specifically for that family. The family chooses from several different mentoring categories, such as personal development, strengthening key relationships, financial wisdom, vocational skills, and educational advancement.

In addition to the mentor teams, Faith Partners also operates a 30-day Crossroads class for new TANF clients in El Paso County, which is held at DHS. This weekly class sets the stage for future successful work participation by encouraging clients to think about their opportunities, changing attitudes about work, and laying the groundwork for expectations of success. In April 2008, the Crossroads class became mandatory for all work-ready TANF clients in El Paso County. Faith Partners developed the Crossroads curriculum after attending Dr. Ruby Payne’s “Bridges Out of Poverty” training, a class that helps participants understand their life-skills and workplace needs.

Results and Lessons Learned

Because Faith Partners has been colocated in the DHS, and cohoused with the TANF unit for the past 10 years, they have been able to fully integrate their services. The Colorado DHS’s TANF/Colorado Works Program provides support to both El Paso County DHS and Faith Partners.7 Offering the Crossroads class, which has been in place for 10 months, is a direct result of the success of Faith Partners in helping families and individuals to become self-sufficient. This new class has presented a true win-win situation for DHS and Faith Partners because both entities have a better product, as clients move from one work participation opportunity (i.e., the

69

Crossroads class) to another (possible mentoring by Faith Partners and job assistance and placement with DHS).

In 2007, Faith Partners provided mentoring services to over 30 families and life-skills education to over 360 TANF clients while operating within a budget of approximately $110,000. In the program’s history, over 1,200 volunteers have become mentors and have mentored over 440 families. By Faith Partners colocating with DHS, communication and information sharing has been facilitated. The ease of sharing has encouraged joint problem-solving and a unique perspective on each partner’s needs and business practices.

Site Four: Pathways, Petersburg, Virginia

Founded in 1995 as an interfaith community development corporation in Petersburg, Pathways, formerly known as Petersburg Urban Ministries, has participation from the Jewish, Muslim, Christian, and Bahia faith communities. Pathways is highly visible in its community, serving a predominantly low-income, African-American population of young people aged 15 to 24. It is also an AmeriCorps member and an affiliate of YouthBuild USA.8 Pathways partners with the local Department of Social Services (DSS), the local Workforce Investment Board (WIB), and other groups to provide job training opportunities, job placement, and retention support for a number of young adults receiving TANF. By offering a holistic approach, Pathways uses an asset-based approach, seeing the full potential of all their clients, and helping them to attain it.

Pathways is unique among the profiles for its interfaith work, which has breadth and depth into many different faith traditions. With 34 congregations and 23 faith groups contributing to their joint mission, Pathways offers a nonsectarian spirituality in its approach to programming, reflecting the interfaith nature of the collaboration. With so many faith groups represented, the volunteers and staff refrain from proselytizing, but may speak about their faith convictions. In general, Pathways emphasizes spiritual attributes, like stewardship of resources and gratitude, on a daily basis. The interfaith nature of the organization is demonstrated in its board of directors and its volunteers. Faith groups are involved in many aspects of volunteering, from working alongside AmeriCorps program participants to tutoring on a weekly basis.

Partnership with TANF Agency and Services to TANF Clients

Pathways has a strong but informal partnership with their local TANF agency: they have a letter of support from DSS that they will refer TANF recipients to Pathways programs, with specific commitments from each party. Pathways’ social workers often work on behalf of TANF-eligible clients, interfacing with their caseworkers to provide information and support. These social workers also help those who qualify for TANF to apply.

70

While not exclusively focusing on TANF recipients, Pathways serves TANF clients in its YouthBuild Program, the AmeriCorps program, and in workforce development through the local WIB. The clients receive a range of supports, including educational classes, tutoring, and job training. They are also engaged in leadership development, community action projects, mentoring, and job shadowing. Specifically, the low-income young people in YouthBuild programs work toward their high-school diploma or General Equivalency Diploma (GED), learn job skills, and contribute to their communities by building affordable housing, thereby changing their own lives as well as benefiting their community. Most of these individuals move from being TANF-eligible to being ineligible due to income earnings after completion of the program. In addition, Pathways is dedicated to providing opportunities for educational advancement to youths and adults, as well as faith support, if the clients express interest. In their personal assessment on entrance, clients have the opportunity to talk about their own faith or spirituality and may follow up on that with their counselor or other staff.

Results and Lessons Learned

Serving 500 youths a year, Pathways also serves 100 TANF clients and referrals in the course of a year. Moreover, 75% to 80% of clients receive a recognized credential in their work of choice. Pathways has a 75% program completion rate, with a 90% or better placement rate in higher education and/or employment after clients graduate from the program.

As an interfaith organization, Pathways is unique among the programs profiled in this study. Instead of finding their differences to be a roadblock, Pathways members have found that when people work together on a project and have a common goal, the participants are able to collaborate more easily. In this case, housing and youths were the two issues that the community agreed needed their intervention, so Pathways began to work with vulnerable youth populations and on community housing issues. Pathways attributes their lack of challenges to the commonly held vision and mission, as well as the inclusive approach they have used from the start of their work. Inspired by their faith backgrounds, the 23 different faith groups were (and continue to be) brought to the table for a project that they all see as important.

Site Five: Seedco, New York, New York

Seedco’s mission as a national nonprofit organization is to work with local partners to create economic opportunities for disadvantaged job seekers, workers, and neighborhood entrepreneurs by implementing innovative, multipartner initiatives that foster financial advancement and security. Over the past decade, Seedco has developed, refined, and replicated an approach for delivering high-quality welfare-to-work services to TANF clients on a large scale by partnering

71

with networks of FBCOs. The FBCOs provide preemployment support and employment and retention services to residents in their communities. While Seedco works in other locations, the focus here is on its work in New York City, with its 14 FBCO partners there.

Structurally, Seedco serves as a management services intermediary, whereby it receives performance-based contracts from state and local TANF agencies and then subcontracts with FBCO partners that are responsible for providing direct services to TANF recipients and applicants. In addition to contract management and interface with TANF agencies, Seedco also provides common program models and tools, centralized management information systems, resource development, technical assistance, and administrative support. FBCOs, in turn, leverage their capacity to engage individuals with multiple employment barriers and provide responsive, high-quality services to them. Because of this cooperative program design, all FBCOs and Seedco share a stake in overall performance, and FBCOs have the opportunity to learn from each other.

In New York City, Seedco’s 14 FBCO partners include a range of organizations, from small community development corporations that develop housing in their communities to larger agencies that offer a range of services to a targeted neighborhood or geographic area. While both FBCOs are part of this network, all of the services provided are secular. The Seedco model allows FBCOs to compete as welfare-to-work service providers because participating FBCOs collectively can reach a level of scale comparable to large providers and the FBCOs benefit from Seedco providing a shared infrastructure and administrative support for the network. By providing the infrastructure elements to its FBCO partners, Seedco is able to leverage the FBCOs’ individual strengths at working with low-income workers. Simultaneously, Seedco pools their resources in a common model that has a far greater reach than the individual FBCOs would be able to achieve, both in terms of numbers of clients served, as well as professional systems to serve and track the clients.9 Seedco identifies challenges faced by multiple partners and areas in need of improvement, and responds accordingly with training, technical support, and other tools.10

Partnership with TANF Agency and Services to TANF Clients

Although Seedco often leverages private foundation dollars to fund service enhancements and innovations, contracts with FBCOs are funded through TANF and Workforce Investment Act (WIA) contracts. Seedco’s model was fully implemented with local TANF funds in 2000. In New York City, the local TANF agency assigns TANF recipients to Job Centers for welfare-to-work services based on where the recipient resides. Once at one of four Seedco-run Job Centers,

72

Seedco staff assign clients to an FBCO partner based on the FBCO’s caseload capacity, geographic proximity, and the preference of the client. In 2008, Seedco will place approximately 8,000 individuals into jobs through the FBCO networks, including 5,000 TANF clients. Overall, Seedco is consistently among the top performers in New York’s welfare-to-work contracting arena in terms of client placement, employment retention, and reduced welfare recidivism.11

Results and Lessons Learned

When Seedco first established its FBCO network, it committed resources to build staff and information technology infrastructure. These front-end investments have resulted in an ability to add new FBCO partners at a negligible cost because they built their internal capacity in the beginning of the FBCO work. With the infrastructure elements already established, Seedco can offer new FBCOs partnership opportunities using those preexisting templates, models, and staff expertise. In addition, Seedco’s centralized back-office support allows the network as a whole to achieve economies of scale. Seedco’s partners collectively share the costs associated with the support they receive, including funder relationship management, contract procurement, data tracking and reporting, and fiscal oversight. And Seedco’s program development department secures new contracts at the city, state, and federal levels as well as with private funders.

COMMON THEMES AND IMPLICATIONS

As discussed in the research overview, findings have not yet coalesced around the features of TANF-FBCO partnerships that make them successful. The larger project of which this Environmental Scan is a part intends to provide specific answers to this question. However, the profiles included here illustrate certain features that seem to be common among most successful programs:

# Location: Each of the sites mentioned the importance of location. For some sites, locating near their government partners allowed for ease of referral and streamlined services. For other sites, a location in a neighborhood or near church partners lent additional credibility. The core message appears to be the importance of a strategic location that resonates with the population targeted for service.

# Volunteer Recruitment: The sites in this study operate different types of programs and, as such, use volunteers differently. However, each of the nonintermediary programs emphasized the importance of finding, recruiting, training, and retaining skilled, appropriate, and motivated volunteers.

# Referrals: The sites, other than Rachel’s House, serve clients that are referred to them from their local TANF offices, and the mechanisms that guide those referrals (e.g. contracts, Memorandums of Understanding [MOUs]) are important.

73

# Targeted Services: All of the sites agreed that services delivered should be targeted to the needs of their clients and capitalize on their own organizational strengths.

# Building Services in Increments: All of the sites agreed about the importance of “starting small” and expanding over time. Specifically, sites described the wisdom of working on smaller contracts first and then pursuing larger opportunities as their infrastructure became more established.

Table 2. Common Themes among Faith-based and Community-based Organizations

Program Location Volunteer

Recruitment Referrals Targeted Services

Incremental Growth

Faith Connections (Wilson, NC) $ $ $ $ $ Rachel’s House (Columbus, OH) $ $ $ $

Faith Partners (Colorado Springs, CO) $ $ $ $ $

Pathways (Petersburg, VA) $ $ $ $ $

Seedco (New York, NY) $ $ $

Table 2 represents the commonality of these themes by program. Location emerged as a key determinant of success. Many of the individual FBCOs indicated that their location in a community was of vital importance, but they differed on what location was best suited for them, which varied based on the services provided and organizational goals. When recruiting volunteers, many of the organizations found that working with other FBCOs greatly enhanced their ability to recruit others interested in serving TANF individuals and families, whether directly, for most of the profiled FBCOs, or indirectly, as in the case of Faith Connections. All of the FBCOs found that targeting their efforts toward specific services greatly enhanced their productivity and maximized the positive impact on clients. Furthermore, all of these FBCOs found that building their services in increments was vital for their overall success. While all the projects have different goals and service offerings, by starting small, the FBCOs were able to learn from their first efforts and adjust their strategies for improving service delivery, growing their capacity, and offering increasingly comprehensive supports to TANF clients.

For client referrals, having a close, formal working relationship with TANF agencies was crucial for program success and growth for Faith Connections, Seedco, and Faith Partners. With a letter of support from the local DSS, Pathways also receives referrals, in an informal partnership with their TANF agency. Because Rachel’s House works with women in prison before they leave, the referral process does not apply to their organization in the same way; however, they do receive TANF funding to serve those who are eligible. Lastly, as an intermediary-supported network of

74

FBCOs implementing a common program model, Seedco differs from the other profiled programs in that this model allows small FBCOs to collectively serve large numbers of TANF referrals. The intermediary model, as shown by the success of Seedco, demonstrates that small FBCOs, operating as a network that is supported by an intermediary, can collectively reach a level of scale comparable to large for-profit and nonprofit providers and successfully compete with these groups.

Implications for Future Research

As noted earlier, this Environmental Scan is the first step in a larger research effort. The findings presented here should be considered preliminary and are not necessarily applicable to all TANF-FBCO partnerships. Future research, both as part of this project and conducted by others, can further examine these relationships.

By looking at a broad range of FBCOs and considering a group with diverse characteristics and geography, the findings from this TANF Initiative will be helpful and informative for other agencies and programs within the federal government seeking to partner with FBCOs to effect real change in the lives of citizens. The criteria for assessing FBCOs could be adapted for future use by other departments and agencies in the areas of research, policy, and practice. It is hoped that this initiative will be used as a promising practice model across the government, as a comprehensive look at “what works” with FBCO partnerships and evaluation. Preliminary findings, such as the apparent importance of starting small, building programs in increments, and targeting services will be explored in the future phases of this research project and tested in the pilot phase. The descriptive profiles included in the Compendium can be illustrative for programs seeking new or enhanced partnerships with FBCO social service providers, and the toolkit will provide guidance for making these partnerships successful. Together these pieces lay a solid foundation for rigorous evaluation of the quality of selected partnerships and additional examination of how these partnerships change and develop over time.

ACKNOWLEDGMENT

This project is funded by the Office of Family Assistance, Department of Health and Human Services Contract No. GS23F8062H, Task Order HHSP233200700227G to ICF International.

NOTES

1. The Baylor Institute for Studies of Religion (ISR) examined the success of faith-based and community initiatives in Ohio, providing numerous insights and highlighting the fact that exemplary programs were able to dramatically increase the number of clients that they

75

served and the services they offered. Similarly, in 2004, Compassion Capital Fund (CCF) funded Caliber Associates (now ICF) to evaluate the Kairos Horizon Program, a faith-based residential rehabilitation program for prisoners and their families. Several of the outcomes investigated suggested that the Kairos Horizon program positively impacted prerelease measures of prison safety and postrelease measures of recidivism (e.g., delaying the onset of rearrest) and support of dependent children.

2. Readiness assessments do not examine the capacity of a given site to participate in model testing but rather their capacity for participating in data collection efforts. Factors may include staff availability, infrastructure, longevity of service, or other considerations.

3. Faith Connections’ assistance funds come directly from churches and individuals in the community, which are used to meet the assistance needs. The only other funding that is necessary is the salary for the Faith Connections’ Program Coordinator, who is a DSS employee. That salary is partially funded by the county and partially by TANF. The funds that come from the churches are deposited into a DSS donation fund account and checks are drawn on this account through the DSS fiscal department. The Faith Connections’ Program Coordinator is responsible for balancing this account.

4. TANF funds in this grant (Strengthening Ohio Families Initiative) are specified for nonreligious salaries: case management (LSW), volunteer coordination, project management, and one third occupancy costs.

5. Ohio has opted out of the federal ban on providing TANF-funded services to drug felons.

6. Five organizations have used the Rachel’s House program model to start similar housing programs for ex-offenders, including the Ohio Department of Rehabilitations and Corrections, Sarah’s Place, House of Bethel, Divine Mercy House, and Lydia’s House.

7. Colorado is a state-supervised and county-administered system under TANF. In order for partnerships—such as that of El Paso County DHS and Faith Partners to work well—policy makers, community leaders, and other bodies of leadership must encourage and support them.

8. YouthBuild is a youth and community development program that simultaneously addresses key issues facing low-income communities, such as housing, education, employment, crime prevention, and leadership development.

9. Monthly program design meetings are typically data-driven meetings. Seedco prepares and distributes a management report that details the performance of the network as a whole as well as the performance of individual FBCOs in meeting placement and retention milestones. The report and the data are a jumping-off point for a more qualitative discussion, with high-performing FBCOs sharing their strategies and practices, and other FBCOs discussing challenges and brainstorming with the group to develop course corrections.

10. This structure also provides economies of scale with respect to training delivered by the intermediary. Seedco can provide training and technical assistance to the network at large because each partner is executing the same programmatic activities. By bringing the network partners together, FBCOs can learn from each other by directly comparing experiences and sharing challenges and successes.

76

11. For the most recently completed TANF welfare-to-work contracts for New York City (which ended in 2006), Seedco ranked first in the city for both its clients’ low welfare recidivism rates and their average wages.

APPENDIX A

The TANF Faith-based and Community Organization Initiative includes four discrete research tasks:

# Develop a Research Compendium: This task aims to understand the state of the field in terms of TANF-FBCO collaboration, document promising practices for FBCOs and intermediary/cooperative entities, and inform the selection of promising practices for inclusion in the study.

# Design a Model Program Pilot: The pilot task will develop a workable strategy for faith-based/employment partnerships, provide guidance for bridging the gap between TANF programs and FBCOs so that they may partner successfully, and isolate the elements of successful partnerships that may be replicated in other sites.

# Provide Site Technical Assistance: This task will provide technical assistance to selected jurisdictions on program model implementation and general operations, test the model elements, and assess the fidelity with which jurisdictions are operating the model.

# Develop a Toolkit: The capstone product of this research effort, the toolkit will highlight replication guidance and advance the field with respect to TANF-FBCO partnerships. The toolkit is intended to facilitate the dissemination of project findings to a broad audience, provide replication advice, and provide usable tools to the field.

Figure A-1 Compendium Framework

EnvironmentalScan

Components:

• Literature review• Field contacts &

extant data• Current work

Product:

• Database

Promising Practices

Components:

• Criteria• Variable of

importance• Data verification calls

Product:

• List for sorting along different dimensions

ReadinessAssessment

Components:

• Site visits• Teleconferences

Product:

• Readiness summaries

Synthesis andReporting

Components:

• Summary of practices• Summary of readiness

assessments

Product:

• Synthesis report• Oral briefing

Compendium Framework

COMPENDIUM250 FBCOs 25-40 FBCOs ~25 FBCOs

EnvironmentalScan

Components:

• Literature review• Field contacts &

extant data• Current work

Product:

• Database

EnvironmentalScan

Components:

• Literature review• Field contacts &

extant data• Current work

Product:

• Database

Promising Practices

Components:

• Criteria• Variable of

importance• Data verification calls

Product:

• List for sorting along different dimensions

Promising Practices

Components:

• Criteria• Variable of

importance• Data verification calls

Product:

• List for sorting along different dimensions

ReadinessAssessment

Components:

• Site visits• Teleconferences

Product:

• Readiness summaries

ReadinessAssessment

Components:

• Site visits• Teleconferences

Product:

• Readiness summaries

Synthesis andReporting

Components:

• Summary of practices• Summary of readiness

assessments

Product:

• Synthesis report• Oral briefing

Synthesis andReporting

Components:

• Summary of practices• Summary of readiness

assessments

Product:

• Synthesis report• Oral briefing

Compendium Framework

COMPENDIUM250 FBCOs250 FBCOs 25-40 FBCOs25-40 FBCOs ~25 FBCOs~25 FBCOs

77

REFERENCES

Archambault, C., Kakuska, C., & Munford, R. (2001). Faith-based partnerships: Charitable choice and state TANF programs. Fairfax, VA: Caliber Associates.

Banks, D., Hercik, J., & Lewis, R. (2004). Rediscovering compassion: An evaluation of Kairos Horizon communities in prison. Fairfax, VA: Caliber Associates.

Carlson-Thies, S. (2000). Charitable choice for welfare & community services: An implementation guide for state, local, and federal officials. Retrieved March 18, 2008, from http://www.urbanministry.org/files/Charitable-Choice_Implementation-Guide.pdf

Deschenes, E.P., Owen, B., & Crow, J. (2007). Recidivism among female prisoners: Secondary analysis of the 1994 bureau of justice statistics recidivism data set. Retrieved March 19, 2008, from http://www.ncjrs.gov/pdffiles1/nij/grants/216950.pdf

Ericson, N. (2001). Public/private ventures’ evaluation of faith-based programs: OJJDP fact sheet. Retrieved March 3, 2008, from http://www.ncjrs.gov/pdffiles1/ojjdp/fs200138.pdf

Green, J. (2007). American congregations and social service programs: Results of a survey. Albany: Roundtable on Religion and Social Policy.

Hercik, J., & Lewis, R. (2003). Building on the promise: Providing training and technical assistance for small non-profit organizations on the federal grants application process. Fairfax, VA: Caliber Associates.

Kennedy, S. S., & Bielefeld, W. (2003). Charitable Choice: First results from three states. Indianapolis, IN: Center for Urban Policy and the Environment.

Monsma, S. (2002). Working faith: How religious organizations provide welfare-to-work services. Philadelphia: Center for Research on Religion and Urban Civil Society, University of Pennsylvania.

Reingold, D. A, Pirogl, M., & Brady, D. (2007, June) Empirical evidence on faith-based organizations in an era of welfare reform. Social Service Review: University of Chicago Press.

Sherman, A. L. (2000). A survey of church-government anti-poverty partnerships. American Enterprise, 11(4), 32–33.

Schneider, J. A. (2006). Social capital and welfare reform: Organizations, congregations, and communities. New York: Columbia University Press.

78

ACCESSIBILITY AND STABILITY OF NONPROFIT SERVICE PROVIDERS: FAITH-BASED AND COMMUNITY-BASED

ORGANIZATIONS IN URBAN AND RURAL AMERICA

Scott W. Allard

Over the past four decades, social services promoting work activity and greater personal well-being—such as job training, adult education, child care, substance abuse or mental health services, emergency assistance—have become a central component of the safety net that assists low-income families in America.1 Until the late 1960s and early 1970s, welfare cash assistance was a primary method for helping poor persons, particularly poor single mothers. Today, however, federal and state spending on welfare cash assistance totals about $11 billion annually, whereas the Congressional Research Service (2003) estimates that means-tested federal, state, and local social service programs receive at least $110 billion in funding each year.2 Contrary to popular perceptions that view welfare cash assistance as the dominant approach to antipoverty assistance, the safety net is composed largely of social service programs that help poor populations overcome barriers to employment and achieve better work outcomes.

The role of social service programs in the public safety net has numerous implications for policy and communities. First, secular and faith-based nonprofit organizations (often referred to as religious nonprofits) typically deliver publicly funded social service programs. Discussing the modern safety net, Smith (2002) concluded that “nonprofit social service agencies have a more central role in society’s response to social problems than ever before” (p. 150). Similarly, DiIulio (2004) stated that “faith-based programs, especially in urban communities, are the backbone of broader networks of voluntary organizations that benefit the least, the last, and the lost of society” (p. 82). Not only does the nonprofit sector administer many government-supported services, it also offers assistance financed through philanthropic and charitable giving. Nonprofit service agencies may be more trusted in high-poverty neighborhoods and more responsive to community priorities than government agencies (Owens & Smith, 2005). As a result, efforts to strengthen faith-based service organizations (FBOs) and community-based nonprofit service providers have become prominent components of national, state, and local antipoverty policy agendas.

Second, delivery of social service programs is very different from delivery of cash assistance programs. While welfare or food stamp benefits can be delivered directly to recipients through the mail or an electronic benefits transfer (EBT) card, most social services cannot be mailed or

79

delivered directly to an individual at home. Instead, clients typically visit a social service agency, often several times, to receive assistance or complete a program. Poor persons who do not live in proximity to relevant service providers may find it difficult to address basic household needs, barriers to employment, or more serious health issues because of limited access to transportation resources and complex commutes.

Finally, because social service funding can vary from year to year, it is important to pay particular attention to the stability of nonprofit community service organizations. Whether caused by changing needs or public priorities, the allocation of service program funds can change from year to year. Public and private funding of social services also is cyclical, typically contracting during periods of economic downturn and tight budgets when the need for assistance often rises. Therefore, for a variety of reasons, nonprofit service organizations must cope continuously with lost or shifting revenue streams. As critical as service accessibility may be to improving outcomes among working poor families, the consistency and stability of providers is also essential to adequately assist low-income populations.

With these considerations in mind, this paper examines several important questions about the role of faith-based and secular community-based nonprofit service organizations:

# Are certain types of faith-based or secular nonprofit service organizations more accessible to poor populations than others?

# How are FBOs and secular nonprofit service organizations funded?

# Is service provision more stable and consistent across FBOs than secular organizations?

To help answer these questions, data from a unique survey of nearly 1,200 faith-based and secular nonprofit social service agencies operating across seven urban and rural sites were analyzed. Overall, the results show FBOs that integrate religious elements into service delivery and secular nonprofit organizations are more accessible to poor populations than FBOs that do not integrate religious elements into service provision. Moreover, these data indicate that a large percentage of FBOs and secular nonprofit organizations experience funding volatility and inconsistency in service provision each year. In addition, the paper draws implications for policies that may strengthen community-based nonprofits and future research into the role these organizations play within the contemporary safety net.

80

PLACE, STABILITY, AND SOCIAL SERVICE PROVISION

The term “social safety net” describes the assistance that seeks to prevent adults and children from falling below a minimum material standard of living. Today’s safety net comprises a bundle of governmental and nongovernmental antipoverty programs targeting low-income populations who lack adequate income, food, housing, or access to health care. The most salient safety net programs are those governmental assistance programs designed to reduce material poverty or address health care needs. Public programs, such as food stamps, welfare cash assistance through the Temporary Assistance for Needy Families (TANF) program, and the Earned Income Tax Credit (EITC) seek to increase poor families’ income and resources, whereas the Medicaid program provides health insurance coverage for low-income elderly populations and working poor families.

Less well understood, however, is the importance of social service programs to the modern American public safety net. Social services are critical avenues through which poor and near-poor households address immediate needs, overcome obstacles to employment, and seek better work opportunities (Edin & Lein, 1998, Gutiérrez-Mayka & Bernd, 2006). Assistance comes through many different programs: substance abuse or mental health, food pantries or soup kitchens, temporary cash or food, child care, job training and adult education, housing, and transportation. The Congressional Research Service (2003) estimates government spending on many different means-tested social service, job training, housing, adult education, and energy assistance programs has doubled in real dollars over the past 30 years. Because nonprofits administer many of these programs, expanded public funding has led to growth of the nonprofit service sector. Salamon (2002) showed that the number of nonprofit human service organizations increased by 115% between 1977 and 1997, with total revenues for those organizations more than doubling during that time. The number of nonprofit human service and job training service providers has continued to grow, increasing by more than 60% between 1990 and 2003. Total revenues for these organizations now reach about $80 billion (Allard, 2008).3 Combining public and private expenditures, it can be estimated that the United States allocates between $150 and $200 billion to social service programs annually.

One implication of maintaining a safety net reliant on social services is that issues of provider accessibility become paramount. For example, a poor person cannot readily participate in a social service program located many miles away, making proximity to service providers critical to receiving help. Information about the services available is likely to be a function of proximity to providers because an individual is more likely to know about the agencies present in their

81

immediate neighborhood than in neighborhoods farther away. Caseworkers often will provide low-income individuals with information about programs in the immediate community. Living in closer proximity to providers will reduce the commuting burden, especially if office visits must be coordinated with already complex trips between home, child care, and work. Further, the limitations of public transportation in many communities and low rates of automobile ownership among low-income households make it even more critical that providers are located near poor populations.4

Compounding the challenge of ensuring adequate access to programs, social service providers often experience fluctuations in the level and sources of program revenues from year to year. Changes in funding are due in part to shifts in public and private priorities that reflect evolving demographic patterns, community needs, or policy agendas. Government agencies and nonprofit philanthropies often allocate their finite resources to reflect moving priorities. As noted earlier, public and private funding for social service programs decrease during economic downturns, when revenues, endowments, and private giving decline. This responsiveness to the economic cycle means that funding available for social service programming most often decreases at the same time that the need for assistance increases.

Because the safety net is financed in this manner, social service organizations devote substantial energy to maintaining program funding, seeking new sources of funds or looking for revenues to replace lost funding sources. Agencies and organizations that cannot receive a consistent flow of revenue or program resources will be forced to cut staff, reduce available services, and/or limit the number of people served. In extreme cases, agencies or organizations may be forced to temporarily close or even permanently shut their doors because of insufficient or inconsistent funding. Volatility in program funding streams not only makes assistance less available to those in need, it destabilizes the agencies and organizations on which the safety net is founded. Understanding how funding is allocated within communities and across agencies, therefore, is critical to identifying where the safety net is most vulnerable.5

THE MULTI-CITY SURVEY AND RURAL SURVEY OF SOCIAL SERVICE PROVIDERS

Despite the importance of accessibility and stability to the success of social service programs, there is relatively little information comparing the accessibility and stability of FBOs with secular nonprofit service organizations.6 To better understand how faith-based service organizations and secular nonprofit service organizations operate within local safety nets, data were analyzed from the Multi-City Survey of Social Service Providers (MSSSP) and the Rural Survey of Social Service Providers (RSSSP), which were completed with executives and

82

managers in three metropolitan areas (Chicago, Los Angeles, and Washington, D.C.) and four multiple county rural sites (southeastern Kentucky, south-central Georgia, southeastern New Mexico, and the border counties of Oregon-California) between November 2004 and June 2006. Organizations self-identified as either governmental, secular nonprofit, or religious nonprofit organizations; the latter are referred to as faith-based organizations in this paper. Each survey then gathered detailed information on location, services provided, clients served, funding, and organizational characteristics from these public and nonprofit service providers. With response rates that exceed 60% in each site, these surveys are the most unique, comprehensive, and geographically sensitive data about social service provision currently available.

As expected, nonprofit service organizations are critical components of safety nets in urban and rural areas. Among the organizations interviewed by the MSSSP, 70% self-identified as secular or faith-based nonprofit organizations and are included in the present analyses; 61% of providers in the RSSSP reported either secular or faith-based nonprofit status. Totaled across the two surveys, there are 1,172 secular or faith-based nonprofit organizations, of which roughly 60% self-identified as secular nonprofits (67% in the MSSSP, 58% in the RSSSP). Survey questions about the frequency with which religious elements are incorporated into service delivery are used to sort FBOs into one of two categories: faith-integrated agency or faith-segmented agency. Faith-integrated agencies are those that report frequent involvement of prayer with clients, promotion of particular religious viewpoints, or discussion of behavioral or lifestyle issues using religious principles in the course of service delivery. These organizations often are small church-based food pantries or emergency assistance programs that help several dozen people each month, but they also include larger organizations that help up to several hundred clients at any point in time with a wide range of material, employment, and personal needs.

Faith-segmented organizations do not frequently incorporate prayer, religious viewpoints, or religious principles into service delivery.7 A typical example of a faith-segmented organization is a local site of a larger agency like Catholic Charities or Lutheran Social Services whose origins are in a faith community, but where faith elements are not actively incorporated into service provision. Other examples of faith-segmented organizations include local nonprofits spun off from places of worship or ministerial associations with the intention of separating the social mission from the religious mission of the organization.

Using this approach, 70% of FBOs in the MSSSP are classified as faith-segmented organizations and 30% are classified as faith-integrated. Rural FBOs were more closely split, as slightly more than 53% of FBOs in the RSSSP are categorized as faith-segmented organizations and 47% are

83

categorized as faith-integrated. See Appendix A for more detail about these two surveys and the questions used to determine religious or secular status.

COMPARING THE FEATURES OF FAITH-BASED AND SECULAR SERVICE PROVIDERS

Table 1 (see Appendix B) examines whether the types of services, organization budget, and percentage of clients living below the poverty line vary among secular nonprofit organizations and FBOs. Consistent with existing research, the top panel of Table 1 indicates that FBOs in urban and rural areas are more likely to offer services to address immediate material needs, such as emergency food or cash assistance, rather than services requiring trained professional staff, such as outpatient mental health treatment, substance abuse programs, or employment-related services.8 For example, faith-integrated organizations are considerably more likely to offer emergency assistance (88% in the MSSSP and 85% in the RSSSP) than secular nonprofit organizations (50% in both the MSSSP and RSSSP).

Also, the degree to which FBOs and secular nonprofit organizations deliver programs intended to improve nonmaterial well-being and work outcomes varies. About half of all secular nonprofit organizations in the MSSSP offer mental health or substance abuse programs; nearly 60% administer employment-related programs. By comparison, only about one third of FBOs in the MSSSP and the RSSSP offer outpatient mental health or substance abuse services. Slightly larger percentages of faith-integrated and faith-segmented organizations offer employment-related services, such as job training or adult education, but again at rates well below secular nonprofit organizations.

Part of the reason that FBOs may have a different programmatic orientation than secular nonprofits is because many faith-integrated and faith-segmented organizations have fewer resources than secular providers. Without funds for full-time professional staff or program resources, FBOs may not be able to offer mental health or employment services. Whereas 51% of secular nonprofit organizations in the MSSSP and 34% in the RSSSP have annual budgets above $1 million (see the middle panel of Table 1), less than one third of FBOs in the MSSSP and less than 10% of all FBOs in the RSSSP report budgets over $1 million. Almost one third of faith-integrated providers in the MSSSP and nearly half in the RSSSP have operating budgets of less than $50,000 annually.

Given that secular nonprofit organizations are more likely than FBOs to provide mental health, substance abuse, and employment-related programs that are utilized by a broad range of poor and

84

nonpoor persons alike, it might be expected that these nonprofits maintain client caseloads containing fewer poor persons than FBOs, which focus primarily on the material needs of the poor. However, this is not the case. As shown in the bottom panel of Table 1, both faith-based and secular nonprofit organizations target most of their programs to populations below the poverty line. Although FBOs, particularly those in rural areas, serve poor persons almost exclusively, nonprofit providers of all types in the MSSSP and RSSSP have caseloads predominantly composed of persons with income below the federal poverty line.

ACCESSIBILITY OF FAITH-BASED VERSUS SECULAR PROVIDERS

Differences in client characteristics are suggestive, but alone they provide little information about the accessibility of social service opportunities in a particular community. Although data from the MSSSP indicate that 63% of nonprofit organizations draw a majority of their clients from within three miles, it is still a concern where nonprofit organizations choose to locate in communities.9 There are a number of factors that shape an agency’s location decision: availability of suitable office space, affordability of space, a mission to serve certain neighborhoods or population groups, access to private donors and other revenue streams, and proximity to adequate densities of potential clients.

To better reflect whether providers locate near high-poverty areas, it is helpful to calculate service accessibility scores in the three urban sites to reflect each residential census tract’s relative access to a particular type of nonprofit organization (faith-integrated, faith-segmented, secular nonprofit) offering basic needs, mental health or substance abuse, or employment-related services to low-income populations. These scores weight for the number of clients served within three miles of a given tract and by the number of poor persons within three miles to control for potential demand. More detail about the construction of the service accessibility measures used is provided in the Appendix A.

Service accessibility scores indicate whether a particular type of service provider is located closer to concentrations of poor populations within a given community. These service accessibility scores can be used to compare service provision across different types of census tracts or neighborhoods. Scores above 1 indicate greater access to service opportunities compared with the average tract or neighborhood. For example, Neighborhood A, with an access score of 1.10 for faith-integrated providers, is located within three miles of 10% more service opportunities delivered by faith-integrated providers than the metropolitan mean tract controlling for supply and demand. If Neighborhood B has an access score of 0.90 for faith-integrated

85

providers, it can be said to be located near 10% fewer service opportunities than the metropolitan mean tract. Also, it can be said that Neighborhood A has access to 22% more service opportunities than Neighborhood B (1.10 ÷ 0.90=1.22). Accessibility scores indicate how the volume of clients served by particular types of nonprofit service providers is distributed across low-poverty versus high-poverty neighborhoods, but they do not indicate whether the supply of services is adequate to meet need in any given neighborhood.10 With these caveats in mind, Table 2 (see Appendix B) reports mean accessibility scores across census tracts with low poverty (poverty rate less than 10%), moderate poverty (poverty rate between 11% and 20%), high poverty (poverty rate between 21% and 40%), and extremely high poverty (poverty rate over 40%).

There is consistent evidence that neighborhoods with higher poverty rates have greater access to secular nonprofit and faith-integrated providers than to faith-segmented service providers. The top panel in Table 2 reports the mean scores for access to emergency cash or food assistance. Low-poverty neighborhoods have nearly twice as much access to faith-segmented providers offering emergency assistance as high-poverty or extremely high-poverty neighborhoods (1.27 versus 0.72 and 0.67, respectively). In contrast, high-poverty and extremely high-poverty areas have greater access to faith-integrated service providers offering emergency assistance than the average neighborhood (1.12 and 1.25, respectively), and access to secular nonprofits offering temporary help with material needs is comparable to the average neighborhood in each city (1.08 and 1.04, respectively).

Similar patterns are evident in the next two panels of Table 2, which report access to outpatient mental health/substance abuse services and to employment-related services. Although mental health and substance abuse services are more resource intensive and less common among FBOs, higher poverty communities have greater access to faith-integrated agencies offering outpatient mental health and/or substance abuse programs. In fact, high-poverty and extremely high-poverty neighborhoods have access to many more mental health and substance abuse service opportunities delivered through faith-integrated organizations than through faith-segmented organizations. Persons living in high-poverty and extremely high-poverty tracts also have higher than average access to secular nonprofit organizations offering these types of services, with access scores about 11% to 12% above the metropolitan mean. Gaps in accessibility between faith-integrated organizations and faith-segmented or secular nonprofit organizations persist when looking at employment-related services. Neighborhoods with poverty rates above 40% have access to nearly twice as many faith-integrated service providers offering employment-

86

related services as faith-segmented or secular nonprofit organizations (1.75 versus 0.94 and 0.85, respectively).

Looking across three different cities and three different types of social services, there is evidence in the MSSSP that faith-integrated service providers are more accessible to residents of high-poverty central city neighborhoods than faith-segmented organizations. Such findings are consistent with expectations that places of worship and religious congregations located in high-poverty communities play a particularly active role in providing assistance to the poor in surrounding communities. Secular nonprofits often have access scores above the metropolitan average, which also highlights the critical role these organizations play in impoverished communities.

It is important to keep in mind, however, that faith-integrated organizations account for just a fraction of the nonprofit service sector and the assistance it provides. In both urban and rural areas, secular and faith-segmented service providers help 10 persons for every person receiving help through a faith-integrated provider. Faith-integrated organizations are well-located with respect to poverty, but very few faith-based or community-based organizations provide enough aid to meet the demand in their surrounding community. Moreover, these results do not capture the intent or motivation of different types of organizations to serve low-income populations. In fact, much of the observed differences in accessibility between faith-integrated organizations and other nonprofits may reflect unique opportunities to lease or utilize office space in high-poverty communities that emerge from close partnerships with places of worship in those communities.

FUNDING THE FAITH-BASED AND SECULAR NONPROFIT COMPONENTS OF THE PUBLIC SAFETY NET

Although the federal government has sought in the past decade to reduce the barriers that FBOs may face in receiving public funding, little data exist to indicate how services are funded across faith-based and secular nonprofit organizations. To address this issue, each survey asks providers whether they receive funding from government grants or contracts, grants or contracts funded by nonprofit organizations or foundations, or from private giving from individuals.11 In addition, organizations are defined as “dependent” on a particular revenue source if they receive more than 50% of total organizational revenues from that source.

The findings presented in Table 3 (Appendix B) indicate that nonprofit organizations receive substantial amounts of government funding. For example, roughly 85% of secular nonprofit organizations in urban and rural areas report receiving government funding of some kind. Close

87

to 60% of those secular nonprofit organizations receiving government funds are dependent on those funds. Such findings are to be expected because the nonprofit service sector operates as a key administrative arm of the expanded public safety net. Thus, without secular nonprofit organizations, many government programs could not be delivered at the street level.

In addition, a sizeable share of FBOs report receiving public funding of some kind. More than half of all faith-segmented providers and a smaller, but substantial, share of faith-integrated providers receive government grants or contracts. On the one hand, these findings run counter to assumptions made by past and current federal initiatives to better connect faith-based nonprofit organizations to public funding opportunities. On the other hand, because the law does not permit FBOs to use public funds to support worship or proselytizing activities or to incorporate faith elements into programs, there may be concern that public funding is supporting programs with explicit religious purposes or intents. Limitations of these survey data, however, should temper these latter reactions and lead to a cautious view of information about public funding of FBOs. Neither the MSSSP nor RSSSP can link receipt of public funds directly to religious activity. In many instances, it is likely that faith-integrated providers fund programs with religious content through nongovernmental revenue sources and use governmental funds for programs without religious content.

Furthermore, Table 3 indicates that governmental funding does not compose a substantial share of operating revenues for faith-integrated organizations. Only 35% of faith-segmented organizations in the MSSSP receiving governmental funding rely on those funds for at least 50% of their total revenues; only 7% of faith-segmented organizations in the RSSSP are dependent on public funds. Less than 20% of faith-integrated organizations that receive public funding are dependent on these funds.12

If governmental grants and contracts are less common among FBOs than secular nonprofit organizations, where do religious nonprofit service providers draw funding from? Most FBOs report revenue from nonprofit organizations or philanthropies and from private donors. Funding from the nonprofit sector and from private donors provides critical support to faith-integrated organizations in urban and rural areas. Among faith-integrated providers in the MSSSP, 56% receive nonprofit grants and over 93% receive private donations. Similar patterns are evident in the RSSSP. While about one third of faith-integrated organizations receiving support from other nonprofit organizations are dependent on those revenue streams for a majority of their funding, more than half of those reporting private donations draw a majority of organizational revenues from those sources. Neither faith-segmented nor secular nonprofit service providers appear to rely heavily on nonprofit and private giving. For example, while three quarters of faith-

88

segmented organizations in the MSSSP receive nonprofit grants and 90% receive private donations, very few of those organizations are dependent on nonprofit grants or private philanthropy for a majority of their operating revenues. This brief snapshot suggests that secular nonprofit organizations are dependent on governmental grants, faith-segmented service providers maintain more balanced funding streams, and faith-integrated organizations are highly reliant on private giving.

Beyond concerns about the sources of program funding, there is also a concern with whether certain types of providers are more vulnerable to funding cuts than others. When looking at reports of decreases in any funding source over the 3 years prior to the survey, both faith-based and secular service providers display fairly high levels of volatility in the composition of agency funding. The bottom panel of Table 3 indicates that secular nonprofit organizations are more vulnerable than FBOs to revenue cuts. Nearly 50% of secular nonprofit organizations in the MSSSP and RSSSP report a decrease in any revenue source in the previous 3 years. By comparison, roughly 40% of faith-segmented organizations and 30% of faith-integrated organizations in these two surveys report a funding decrease. The fact that secular nonprofit organizations are more likely to experience funding cuts may be a reflection of their substantial dependence on public funding sources that change frequently from year to year.

Funding cuts often translate into changes in program offerings, staffing levels, numbers of clients served, or, in the extreme, closure of a facility. Almost 55% of all nonprofit service providers in urban and rural areas report reducing service provision in the previous year as a result of funding cuts. The last four rows of Table 3 report specific programmatic responses to funding cuts. Although secular nonprofit organizations appear slightly more likely to report reductions in operations than FBOs, often these differences are not statistically significant. For instance, 37% of faith-integrated organizations in the MSSSP and 33% in the RSSSP reduced the number of clients served in response to recent funding cuts; 39% of secular nonprofit organizations in the MSSSP and 43% in the RSSSP reported serving fewer clients as a result of funding decreases. Secular nonprofit organizations, probably because they are larger organizations on average and carry larger staffs than FBOs, are more likely to reduce staff in response to funding cuts. Roughly 60% of secular nonprofit organizations in both urban and rural areas indicated staffing reductions in the wake of funding losses, compared with about 40% of faith-segmented organizations. Reflecting the vulnerability associated with being a small and modestly funded organization, faith-based service providers appear more likely to temporarily close because of funding cuts than secular nonprofit organizations. These differences, however, do not reach conventional levels of statistical significance. Although different nonprofit organizations draw

89

on different combinations of funding, it appears that issues of volatility and instability in service delivery arrangements are more the rule than the exception across urban and rural nonprofit service sectors.

ORGANIZATIONAL, POLICY, AND RESEARCH IMPLICATIONS

Faith-based and community-based nonprofit organizations are critical components of the contemporary American public safety net. Data from the MSSSP and RSSSP indicate that faith-based service organizations play a particularly critical role in providing basic material assistance to low-income households, many of whom may not be eligible for governmental assistance. Yet the fact that more than one-third of FBOs in urban and rural areas offer mental health, substance abuse, and employment-related services suggests that they also offer assistance that addresses a broader range of barriers to self-sufficiency among the poor than might otherwise be assumed. Moreover, when looking at the accessibility of secular and faith-based service organizations in the three urban sites, there is evidence that faith-integrated organizations—those that incorporate matters of faith in service provision—are the most geographically accessible sources of support to high-poverty communities across a number of different program areas. Secular nonprofit organizations, which provide more assistance and more services that address barriers to employment than FBOs, are also quite accessible to high-poverty communities.

Apart from issues of accessibility, the nonprofit sector exhibits substantial variability in service delivery because of volatility in revenue fluctuations. Secular nonprofit organizations are particularly vulnerable to funding cutbacks, in part because they rely so heavily on public revenue streams that can be quite responsive to economic downturns or tight budgetary times. Yet anywhere from 25% to 40% of FBOs in urban and rural areas have experienced a recent cut in funding, and as a result many of these organizations were forced to reduce services, clients, or staff.

These data also highlight important distinctions between the characteristics of faith-based and secular nonprofit organizations operating in urban versus rural areas. FBOs and secular nonprofit organizations comprise a smaller share of all social service providers operating in the four rural communities examined by the RSSSP than in the urban communities of the MSSSP. Nonprofit organizations in these four high-poverty rural areas, faith-based and secular nonprofit organizations alike, also have smaller budgets and fewer resources than those located in the three urban areas. Modest resource levels reflect lower levels of wealth in these particularly poor rural areas, which translates into fewer private resources targeted at social services. Rural-urban

90

differences also reflect the lack of public resources available in high-poverty rural areas for social service grants or contracts. Nevertheless, rural service providers work with individuals who are as poor and disadvantaged as those in urban areas. All high-poverty communities are challenged to find adequate resources for programs of assistance, but the data presented here suggest that rural communities face a particularly steep challenge in marshalling adequate resources to meet the needs of the working poor.

Taken together, these findings indicate that initiatives to strengthen faith-based and community-based nonprofit organizations are critical steps in achieving a sound public safety net and increasing the availability of assistance to poor populations. Along with efforts to promote community-based nonprofit organizations, other actions can fortify the safety net. For instance, given the volatility in service delivery reported by both faith-based and secular nonprofit organizations, policy makers and community leaders may seek to ensure that funding for social service programs is more stable and predictable than currently is the case. One step toward ensuring stability is for federal agencies, states, and communities to vigorously maintain public commitments to funding social service programs. As the data here suggest, the nonprofit sector is unlikely to replace substantial cuts in government funding of programs. In addition to maintaining public funding commitments, policy makers and local leaders may focus on helping nonprofit organizations achieve greater diversification within their funding portfolios to better weather lost program funding when it occurs. Moreover, government agencies and community-based nonprofit organizations may wish to pay greater attention to how cuts in social service programs affect impoverished neighborhoods and communities. Such efforts will allow communities to better coordinate programmatic responses to unmet needs.

As critical as it is to ensure funding for local service providers, more attention needs to be paid to the space and facility needs of these organizations. Finding affordable space that is suitable for service provision is a challenge. Even when agencies can locate suitable space they can afford, there may be pushback from local residents or other building tenants to renting space to nonprofit organizations that help poor populations. Declining poverty rates in many central city neighborhoods and increasing poverty rates in nearby suburban communities also pose complications for providers.13 Because most funding is for programs or services and not for relocation or space acquisition, nonprofit organizations will struggle to remain proximate to client populations as poverty moves outward from cities. Therefore, initiatives to support faith-based and community-based nonprofit organizations need to address the space and facility challenges confronting many service providers. Such assistance may come through direct

91

funding for capital investment or through efforts to create office space where nonprofits can colocate to be accessible to poor populations.

Despite the centrality of faith-based and secular social service organizations to local safety nets, there is relatively little information available about these organizations. Rigorous and objective research of nonprofit social service provision, therefore, also will play an important role in identifying how government and communities can best support human service nonprofit organizations. Of great importance is research that evaluates program outcomes among different types of faith-based and secular nonprofit organizations. In addition, research needs to develop better measures for tracking how faith activities matter or shape individual-level outcomes, as well as to understand whether service delivery within FBOs differs from secular nonprofit organizations.

While this paper generates important insight into issues of service delivery, the presence of mismatches between nonprofit organizations and poor persons, program funding, and organizational stability, future research will need to seek stronger measures of program accessibility, the adequacy of service provision relative to need, and program quality. To permit meaningful comparisons across communities, data collection activities will need to be structured to be geographically representative of several different regions or metropolitan areas. Moreover, there is a need for further inquiry into the needs of working poor families and the factors shaping utilization of social service programs to address these needs. It is important for research to assess which strategies best support modestly resourced community-based service organizations and to help those organizations adopt high-quality programs capable of serving large numbers of clients. Along these lines, scholarly work may examine whether public investment in new nonprofit organizations creates more new service opportunities or more accessible service opportunities than strategies that target funding at faith-based and secular nonprofit organizations currently operating.

Combined, such efforts by policy makers, community leaders, and researchers can expand the capacity of faith-based and community-based service providers located in high-poverty communities. Such efforts should translate into more direct assistance available to the working poor and improvement in their well-being.

ACKNOWLEDGMENTS

This project was supported by research grants from the Brookings Institution, Brown University, the Center for Policy Research at Syracuse University, and the Department of Housing and

92

Urban Development, as well as support from the Institute for Research on Poverty at the University of Wisconsin–Madison, the Institute for Policy Research at Northwestern University, and the National Poverty Center at the University of Michigan.

NOTES

1. Social services are defined slightly differently in other research settings (see Smith 2002).

2. Amounts are reported in $2006. However, such an estimate certainly understates the size of the public social service sector, as it excludes a wide range of job training, substance abuse and mental health treatment, child care, housing, and energy assistance programs operated by federal, state, and local governments.

3. Amounts are reported in $2006. Author’s estimates of 501(c)(3) organizations are based on data from the Urban Institute’s National Center for Charitable Statistics, Washington, DC. These estimates include only organizations likely to provide direct services to low-income adults on-site. In addition to excluding advocacy groups, mental health and substance abuse service providers, housing and shelter, and civil rights or legal aid programs are excluded from these calculations because it is not possible to discern which agencies within these categories provide direct services to working age adults on-site or in an outpatient capacity.

4. Existing studies have found evidence of mismatches in the location of social service providers. Allard (2008) showed that neighborhoods where the poverty rate is over 20% have access to almost half as many social service opportunities as neighborhoods where the poverty rate is less than 10%. Grønbjerg and Paarlberg (2001) presented evidence that Indiana communities with higher poverty rates are home to fewer nonprofit service organizations per capita than communities with lower poverty rates. These findings fit with work by Mosley, Katz, Hasenfeld, and Anheier (2003), which showed that high-poverty neighborhoods in South and East Los Angeles are underserved by nonprofit service providers compared with other impoverished areas of Los Angeles County. Likewise, Peck’s (2008) analysis of nonprofit service organizations in Phoenix indicated that providers are less accessible to high-poverty areas near the central city than to low-poverty areas outside the central city.

5. Research has shown that service program funding is volatile. Grønbjerg (2001) and Salamon (1999) found that public funding declined substantially in real dollars from the mid-1970s to the mid-1990s, two decades punctuated by periodic economic downturns and budget crises. Examining the funding of service providers in southern California, Joassart-Marcelli and Wolch (2003) noted that “poor people who reside in the poorest cities of the region are served by nonprofit organizations with lower levels of expenditures, have to share the services of each nonprofit organization with larger numbers of poor people, and hence are likely to receive less and/or lower quality services” (p. 92). Looking at state governments, Johnson, Lav, and Ribeiro (2003), and Smith, Sosin, Jordan, and Hilton (2006) linked recent state fiscal crises to cuts in public expenditures for social service programs. Allard (2008) showed that funding cuts reduce the assistance available to the poor, hamper the effectiveness of programs, and destabilize nonprofit organizations on which the safety net depends.

93

6. Most existing data sources either are unable to make distinctions between different types of faith-based and secular nonprofit organizations, or the data focus exclusively on congregations.

7. To be clear, however, these items do not capture the presence of religious symbols, the degree to which religious elements are embedded within organizational culture, or the specific nature of an agency’s religious affiliations or partnerships. Moreover, nonprofit service organizations receiving public funding may be more likely to downplay their faith connections or activities rather than risk admitting activity that may jeopardize those public funds.

8. Chaves (2002) found that less than 10% of congregations are involved in providing services outside basic food, clothing, or housing needs. Also see DeVita (1999) and Graddy (2006).

9. Respondents to the RSSSP were asked how long it would take the average client to drive to their site. The typical nonprofit provider in the RSSSP, faith-based or secular, indicated the average commuting time by car was 15 minutes.

10. In fact, it should be noted that almost 70% of nonprofit organizations in the MSSSP and RSSSP—FBOs and secular organizations alike—report increases in demand for services in recent years. Access scores do not account for whether programs are of high or low quality. Scores do not speak to how all public and nonprofit resources are allocated across a community or for the length of time in which a client typically participates in a program.

11. Although it is important to make distinctions between types of public revenue, government fee-for-service reimbursements were coded as contract or grant revenue to simplify the requests made of providers during the telephone survey. It is important to note, however, that data on Medicaid fee-for-service reimbursements were collected separately. While about one third of secular nonprofit organizations receive Medicaid funding, less than 10% of all FBOs in the MSSSP and RSSSP receive such funds. In most instances, even when nonprofit organizations receive Medicaid funds, it does not account for a large percentage of operational revenues. As a result, most of the analysis here focuses on other sources of government funding.

12. These findings are consistent with other studies. Monsma (1996) concluded that child service agencies high on his religious practice scale were less likely to be dependent on public funds than secular nonprofit providers or faith-based providers exhibiting low levels of religiosity.

13. See Berube and Kneebone (2006) and Jargowsky (2003).

94

APPENDIX A

Data for this paper were drawn from the Multi-City Survey of Social Service Providers (MSSSP) and the Rural Survey of Social Service Providers (RSSSP), which completed telephone surveys with executives and managers from more than 2,200 governmental and nonprofit social service providers in three cities (Chicago/Cook County, Los Angeles/Los Angeles County, and metropolitan Washington, D.C.) and four high-poverty rural areas between December 2004 and August 2006. MSSSP interviews in metropolitan Washington, D.C., included agencies located in the District of Columbia, as well as Prince George’s County and Montgomery County in Maryland to the northeast and communities in northern Virginia (Alexandria, Arlington, Loudoun County, Fairfax County, and Prince William County). The RSSSP was completed in four multicounty regions: south-central Georgia (Atkinson, Bacon, Ben Hill, Berrien, Coffee, Jeff Davis, Pierce, and Ware Counties); southeastern Kentucky (Bell, Clay, Harlan, Jackson, Knox, Laurel, Rockcastle, and Whitley Counties); southeastern New Mexico (Chaves, Curry, Debaca, Eddy, Lea, and Roosevelt Counties); and an Oregon-California border site composed of 10 counties (Del Norte, Modoc, and Siskiyou Counties in California; Coos, Curry, Douglas, Jackson, Josephine, Klamath, and Lake Counties in Oregon). In addition to questions about services available, faith-based status, and location, the longer surveys in the MSSSP and RSSSP asked respondents more than 100 questions about client characteristics, connections to community organizations, funding streams, and other pertinent organizational characteristics.

Each survey drew respondents from databases of government and nongovernment service agencies constructed for each city or rural region from community directories, social service directories, county agency referral lists, phonebooks, and Internet searches. Churches listed in community directories as providing social services were included in the survey. Providers were contacted by each survey if they operated programs at low or no cost in one of several service areas: welfare-to-work, job training, mental health, substance abuse, adult education, and emergency assistance. The MSSSP began with 5,313 providers, compared with 1,266 in the four rural regions covered by the RSSSP. Verification calls were made to identify agencies that were operational and currently offering services on site to low-income populations. Slightly less than half of all agencies contacted in the MSSSP (2,183 of 5,313) were invited to complete a longer telephone survey; about three quarters of agencies contacted by the RSSSP (964 of 1,266) were eligible for the longer survey. The remaining organizations were either no longer operational, did not provide services at their location, or did not offer programs to low-income persons at low or no cost. The MSSSP completed interviews with 1,487 of the 2,183 agencies eligible for the longer survey (response rate of 68%); the RSSSP completed surveys with 588 of the eligible 964

95

social service providers (response rate of 61%). The poverty rate of the neighborhood in which a provider is located was not statistically related to whether the provider completed the MSSSP. Other organizational characteristics do not appear to have a meaningful impact on the likelihood of response. Similar results were found when examining response rates in the RSSSP.

This paper examines data from faith-based and secular nonprofit organizations interviewed by the MSSSP and RSSSP. Secular or faith-based nonprofit status was determined by answers to the following three questions:

Do you consider your organization to be government, private nonprofit, or private for-profit?

(1) Government (2) Nonprofit (3) For-profit (9) DK/NA

Do you consider your organization to be religious or secular?

(1) Religious (5) Secular (9) DK/NA

Is your organization a religious congregation (i.e., church, synagogue, temple, mosque)?

(1) Yes (5) No (9) DK/NA

The degree of involvement of religious activities in service provision was determined by answers nonprofit organizations provided to the following three questions:

Would you say the following activities occur frequently, occasionally, or not at all at your site?

Staff or volunteers pray with a client.

(1) Frequently (2) Occasionally (3) Not at all (9) DK/NA

Staff or volunteers promote a particular religious viewpoint to a client.

(1) Frequently (2) Occasionally (3) Not at all (9) DK/NA

Staff or volunteers discuss lifestyle or behavioral issues using religious principles.

(1) Frequently (2) Occasionally (3) Not at all (9) DK/NA

Of the 1,304 organizations in the MSSSP that provided information about their public versus nonprofit status and the involvement of faith activities in service provision, 911 (70%) are secular or religious nonprofits and are included in the analyses above. In the RSSSP, 261 of the

96

97

431 organizations (61%) that answered questions about nonprofit status and religious activities are included in this analysis as secular or religious nonprofits. Totaled across the two surveys, 1,172 secular and faith-based nonprofit organizations are included in the analyses reported above.

Based on job accessibility scores calculated previously (see Allard and Danziger 2003; Raphael 1998), city-specific service accessibility scores with data from the MSSSP were calculated as follows. First, the number of clients served by all agencies or a particular type of agency located within three miles of each residential census tract (using tract centroid-to-centroid distances) is totaled. To avoid double-counting, providers were asked to estimate the number of individual clients receiving help and were asked not to double-count clients that may be receiving help from many different programs within an agency. Subsequent site visits to agencies responding to the MSSSP and RSSSP indicate that these estimates are good approximations of supply of services. To account for potential demand for services, the number of individuals with income below the poverty line within three miles of each residential tract were summed. Then the number of clients served was divided by the number of persons in poverty within 3 miles. To be able to compare tracts to each other, this tract-specific access score was divided by the average of that access score for the metropolitan area.

Thus, a set of demand-, distance-, and organization-weighted service accessibility scores was calculated as follows: Ai = Σ(CSi) ÷ Σ(Pi), where Ai is the initial access score for tract i. CSi reflects the number of providers offering a particular service (S) to low-income adults within 3 miles of tract i, multiplied by the number of clients served in each agency in a typical month (C). To account for potential demand, divide by the total number of persons living below the poverty line (Pi) within 3 miles of tract i. To make service accessibility scores more readily interpretable, divide each tract’s score for a given access measure Ai by the metropolitan area mean score for that particular access measure.

Percentage of Service Organizations

MSSSP Providers RSSSP Providers

Faith-

Integrated Faith-

Segmented Secular

Nonprofit Faith-

Integrated Faith-

Segmented Secular

Nonprofit

Emergency Assistance 87.9ab 68.7ac 49.8abc 85.1a 87.0b 50.0ab

Mental Health/Substance Abuse 37.8 32.7a 51.1a 31.9a 9.3ab 36.3b

Employment-related Services 43.3a 51.2 57.6a 38.3 33.3a 56.0a

Annual Budget >$1 million†‡ 34.2 25.3 50.6 4.7 10.4 33.6

Annual Budget $1 million–$200,000 22.4 35.0 33.4 14.0 22.9 25.2

Annual Budget $200,000–$50,000 13.2 26.9 11.5 32.6 33.3 22.1

Annual Budget <$50,000 30.3 12.9 4.5 48.8 33.3 19.1

0–25% Clients in Poverty † 6.9 14.7 9.8 8.7 3.7 7.1

26%–50% Clients in Poverty 9.2 9.3 13.3 4.4 9.3 15.5

51%–75% Clients in Poverty 19.5 23.0 19.8 19.6 9.3 18.7

>75% Clients in Poverty 64.4 52.9 57.1 67.4 77.8 58.7

N 91 211 609 47 54 160

a, b, c Notations identify sets of paired cells within a row associated with a particular survey, where the mean difference in service access between the two cells is significant at the .10 level or below.

Table B-1. Comparing Service Provision across Faith-Based and Secular Service Organizations in the Multi-City Survey of Social Service Providers (MSSSP) and Rural Survey of Social Service Providers (RSSSP)

† Chi-square tests indicate statistically significant variation in this panel at the .10 level or below across faith-integrated, faith-segmented, and secular nonprofit organizations in the MSSSP.

‡ Chi-square tests indicate statistically significant variation in this panel at the .10 level or below across faith-integrated, faith-segmented, and secular nonprofit organizations in the RSSSP.

APPENDIX B

98

99

Table B-2. Access to Faith-Based and Secular Service Organizations in the Multi-City Survey of Social Service Providers (MSSSP)

Mean Service Access Score

Low-Poverty

Tract (Poverty Rate

0%–10%)

Moderate-Poverty Tract(Poverty Rate11%– 20%)

High-Poverty Tract

(Poverty Rate 21%– 40%)

Extremely High- Poverty

Tract (Poverty Rate

>40%)

Mean Access to Emergency Assistance Services Delivered by…

Faith-Integrated Nonprofits † 0.96 0.90 1.12 1.25

Faith-Segmented Nonprofits † 1.27abc 0.84a 0.72b 0.67c

Secular Nonprofits † 0.91 1.08 1.08 1.04

Mean Access to Mental Health and Substance Abuse Services Delivered by…

Faith-Integrated Nonprofits † 0.85ab 0.85cd 1.31ac 1.45bd

Faith-Segmented Nonprofits † 1.49abc 0.77a 0.47b 0.24c

Secular Nonprofits † 0.95a 0.96 1.11a 1.13

Mean Access to Employment-related Services Delivered by…

Faith-Integrated Nonprofits † 0.67abc 0.98ade 1.47bd 1.75ce

Faith-Segmented Nonprofits 0.97 1.13 0.95 0.94

Secular Nonprofits † 1.09 0.90 0.97 0.85

Note: Access scores are weighted to reflect supply of assistance and relative demand for assistance. †=F-tests indicate a statistically significant difference in access to a particular type of provider at the .10 level or below across tract poverty rate.

a, b, c, d, e Notations identify sets of paired cells within each row where the mean difference in service access between the two cells is significant at the .10 level or below.

Sources: Multi-City Survey of Social Service Providers; U. S. Census Bureau 2000.

Percentage of Service Organizations

MSSSP RSSSP

Faith-Integrated

Faith-Segmented

Secular Nonprofit

Faith-Integrated

Faith-Segmented

Secular Nonprofit

Received Government Funding in Previous 3 Years 33.7ab 58.4ac 83.3bc 14.9ab 53.9ac 87.9bc

Dependent on Government Funding 16.7a 34.5b 57.3ab 14.3a 7.4b 62.5ab

Received Nonprofit Funding in Previous 3 Years 55.6ab 74.2a 73.4b 31.9ab 60.4a 55.0b

Dependent on Nonprofit Funding 33.3ab 15.6a 11.4b 28.6 13.3 15.5

Received Private Giving in Previous 3 Years 93.4a 90.0b 74.5ab 95.7a 94.4b 50.0ab

Dependent on Private Giving 52.0ab 17.9ac 5.9bc 73.8ab 52.1ac 5.7bc

Report Decrease in Funding from Any Revenue Source in Previous 3 Years

29.7a 39.3b 49.1ab 25.5a 44.4 47.5a

Reduced Staff in Previous Year due to Funding Decrease

44.4 48.2 61.5 8.3a 29.2b 63.2ab

Reduced Services in Previous Year due to Funding Decrease

44.4 51.8 44.6 41.7 62.5 48.7

Reduced Clients in Previous Year due to Funding Decrease

37.0 40.2 39.3 33.3 50.0 43.4

Temporarily Closed Site in Previous Year due to Funding Decrease

11.1 4.9 7.7 16.7 12.5 7.9

a, b, c Notations identify sets of paired cells within a row associated with a particular survey, where the differences in percentage of providers receiving funding from a particular source, being dependent on that source, or experiencing cutbacks are significant below the .10 level.

Table B-3. Funding and Stability across Faith-Based and Secular Service Organizations in the Multi-City Survey of Social Service Providers (MSSSP) and Rural Survey of Social Service Providers (RSSSP)

Note: Providers are defined as dependent on a particular revenue source if they receive more than 50% of total revenues from that source.

100

REFERENCES

Allard, S. W. (2008). Out of reach: place, poverty, and the new American welfare state. New Haven, CT: Yale University Press.

Allard, S. W., & Danziger, S. (2003). Proximity and opportunity: How residence and race affect the employment of welfare recipients. Housing Policy Debate, 13, 675–700.

Berube, A., & Kneebone, E. (2006). Two steps back: city and suburban poverty trends, 1999–2005. Washington, D.C.: The Brookings Institution, Metropolitan Policy Program, Living Census Series.

Chaves, M. (2002). Religious congregations. In L. M. Salamon (Ed.), The state of nonprofit America. Washington, D.C.: The Brookings Institution Press.

Congressional Research Service. (2003). Cash and noncash benefits for persons with limited income: Eligibility rules, recipient and expenditure data, FY2000–FY2002. Report No. RL32233.

DeVita, C. J. (1999). Nonprofits and devolution: What do we know? In E. T. Boris & C. E. Steuerle (Eds.), Nonprofits and government: Collaboration and conflict. Washington, D.C.: Urban Institute.

DiIulio, J. J., Jr.. (2004). Getting faith-based programs right. Public Interest, 155, 75–88.

Edin, K., & Lein, L. (1998). The private safety net: The role of charitable organizations in the lives of the poor. Housing Policy Debate, 9, 541–573.

Graddy, E. A. (2006). How do they fit? Assessing the role of faith-based organizations in social service provision. Journal of Religion and Spirituality in Social Work, 25, 129–150.

Grønbjerg, K. A. (2001). The U.S nonprofit human service sector: A creeping revolution. Nonprofit and Voluntary Sector Quarterly, 30, 276–297.

Grønbjerg, K. A., & Paarlberg, L. (2001). Community variations in the size and scope of the nonprofit sector: Theory and preliminary findings. Nonprofit and Voluntary Sector Quarterly, 30, 684–706.

Gutiérrez-Mayka, M., & Bernd, E. (2006). How is the region doing? Human service use and service availability in Allegheny County, PA. Pittsburgh: Forbes Funds. Retrieved from http://www.forbesfunds.org/docs/2006Tropman_study2.pdf

Jargowsky, P. A. (2003). Stunning progress, hidden problems: The dramatic decline of concentrated poverty in the 1990s. Washington, D.C.: The Brookings Institution, Metropolitan Policy Program, Living Census Series.

101

Joassart-Marcelli, P., & Wolch, J. R. (2003). The intrametropolitan geography of poverty and the nonprofit sector in Southern California. Nonprofit and Voluntary Sector Quarterly, 32, 70–96.

Johnson, N., Lav, I. J., & Ribeiro, R. (2003). States are making deep budget cuts in response to the fiscal crisis. Washington, D.C.: Center on Budget and Policy Priorities.

Monsma, S. V. (1996). When sacred & secular mix. Lanham, MD: Rowman & Littlefield.

Mosley, J. E., Katz, H., Hasenfeld, Y., & Anheier, H. A. (2003). The challenge of meeting social needs in Los Angeles: Nonprofit human service organizations in a diverse community. Los Angeles: University of California, Los Angeles, School of Public Policy and Social Research, Center for Civil Society.

Owens, M. L., & Smith, R. D. (2005). Congregations in low-income neighborhoods and the implications for social welfare policy research. Nonprofit and Voluntary Sector Quarterly, 34, 316–339.

Peck. L. R. (2008). Do anti-poverty nonprofits locate where people need them? Evidence from a spatial analysis of Phoenix. Nonprofit and Voluntary Sector Quarterly, 37, 138–151.

Raphael, S. (1998). The spatial mismatch hypothesis and black youth joblessness: Evidence from the San Francisco Bay Area. Journal of Urban Economics, 43, 79–111.

Salamon, L. M. (1999). America’s nonprofit sector. New York: Foundation Center.

Salamon, L. M. (2002). The resilient sector: The state of nonprofit America. In L. M. Salamon (Ed.), The state of nonprofit America. Washington, D.C.: The Brookings Institution Press.

Smith, S. R. (2002). Social services. In L. M. Salamon (Ed.), The state of nonprofit America. Washington, D.C.: The Brookings Institution Press.

Smith, S. R., Sosin, M. R., Jordan, L., & Hilton, T. (2006). State fiscal crises and social services. Paper presented at the meeting of the Association for Public Policy Analysis and Management Conference, Madison, WI.

102

THE EMERGING ROLE OF FAITH-BASED ORGANIZATIONS IN THE LOW-INCOME HOUSING MARKET

Richard Hula Cynthia Jackson-Elmoore

Laura Reese

The social, cultural, political, and market dynamics of America’s urban areas have altered the institutional landscape, raising questions about the role of government in service delivery. New service delivery arrangements, particularly through nonprofit organizations, are more widely accepted. Faith-based organizations (FBOs) have become particularly important actors as a result of their traditional roles as producers and distributors of public goods and services. The Bush administration has advocated Charitable Choice—as noted in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA)—as a key component of the Faith-based and Community Initiative, and federal agencies increasingly rely on FBOs to provide a range of services. State governments are adopting similar policies. These fundamental changes in service delivery mechanisms raise important issues about both the effectiveness of the new arrangements (i.e., which services get delivered to what populations) and the potential politicization of FBOs.

Surprisingly little is known, however, about the impact of these changes, as most available information is anecdotal and based on media accounts. Thus, there is clearly a need for a more rigorous, objective, and comprehensive assessment of FBO involvement in service delivery (Hula, Jackson-Elmoore, & Reese, 2007).

Many nonprofit organizations, FBOs in particular, are engaging in a holistic approach to providing housing along with services that address needs that go beyond basic shelter (Cohen, Mulroy, Tull, White, & Crowley, 2004; Housing Plus Services Committee, 2002, 2006). The connection between housing and other social service activities is logical because housing activities build other types of community social capital (Hays, 2002). Faith-based housing activity predates or underpins most other social service provision.

This paper considers the nature of faith-based and secular service delivery models in housing and highlights several issues: (1) the relationships between faith-based and secular nonprofit organizations in providing housing services, (2) the institutions engaged in and the organizational arrangements used for delivering those services, and (3) implications for the

103

future. In addition, it reports the preliminary findings of an ongoing project that explores the role of FBOs in providing housing in inner cities.1

The focus of the current analysis is a comparison of faith-based and secular organizations functioning in the housing arena in Michigan.2 Of particular interest are (1) organizational characteristics, (2) housing output in terms of projects and units, (3) perceived constraints on expanding organizational capacity, and (4) the network of partners and collaborators that the organizations operate within. Overall, the evidence suggests that FBOs are at least as productive as secular housing providers and make good partners for government agencies.3

It is against this backdrop that this analysis is presented. First, the paper discusses in detail the political, market, social, and other contextual factors that have helped to shape the nature of faith-based involvement in urban housing. Key themes from existing research literature are highlighted, as are the study’s main research aims. A brief overview is provided of the study data and methods, followed by a discussion of findings and implications for future research, policy, and practice.

BACKGROUND

The merging of religious and secular organizations in the delivery of social services is not a new phenomenon (de Toqueville, 1969; Hodgkinson & Weitzman, 1993; Netting, 1984; Wineburg, 1992). The most recent iteration of the discussion has roots in the mid-1970s, when there were calls for alternative service delivery mechanisms capable of bridging the gap between a government infrastructure that was sometimes viewed as inefficient, unresponsive, impersonal, and alienating, and the public it serves. By the 1980s, the discussion evolved into the relative value of privatization to encourage private market actors to assume service delivery roles typically reserved for the government (Gormley, 1991; Hula, 1988; Jackson, 1994; National Performance Review, 1993; Savas, 2000). The subsequent and related movement to nonprofit organizations in general and FBOs in particular is an alternative to the standard market model driven by for-profit actors. The reasoning behind this movement is that “mediating institutions” inherent in the nonprofit sector provide a unique promise and possibility not evident in other alternatives. Those mediating institutions most relevant to, and potentially capable of, addressing many pressing social problems include the family, neighborhood, church, and voluntary associations (Berger & Neuhaus, 1977).4

The service delivery role ascribed to FBOs is reinforced in judicial, legislative, and executive decisions. For example, a 1988 U.S. Supreme Court decision affirms that religious organizations

104

can be involved in the delivery of social services (Bowen v. Kendrick, 1988). Some scholars and practitioners have called for government funding of faith-based service delivery to help stabilize the funding in the sector (National Commission on Civic Renewal, 1998; Wineburg, 1993). Congress appears sympathetic to this move, as evidenced in the passage of PRWORA authorizing Charitable Choice as one of the components of the act (P.L. 104-193). The intent of Charitable Choice is to level the playing field for faith-based and community-based organizations to compete for federal funding for service delivery. President Bush’s 2001 executive order establishing the Faith-Based and Community Initiative provides a mechanism to coordinate activities between the public and nonprofit sectors, with a particular emphasis on faith-based and community-based organizations. These actions and others signal government support for an expanded role for FBOs.

The Role of the Nonprofit Sector

The evolution of the role of the nonprofit sector, and specifically FBOs, in housing is particularly relevant to this discussion.5 The involvement of nonprofit organizations in the housing arena dates back to the late 1800s and early 1900s with the establishment of settlement houses and cooperatives for working families (Birch & Gardner, 1981; Keating, Rasey, & Krumholz, 1990). Growing momentum for community-based action set the stage for governmental initiatives to formalize the role of nonprofit organizations in the housing arena. In 1959, the federal Section 202 program provided an opportunity for nonprofit development of affordable housing for disabled and elderly populations. The federal government also established initiatives in the 1960s and 1970s that were supportive of nonprofit organization involvement in housing development.6 Federal, state, and local governments continue to enlist the help of the nonprofit sector to alleviate the country’s burgeoning housing pressures (Bishop, 1991; Bratt, Keyes, Schwartz, &Vidal, 1994; Goetz, 1992; Vidal, 1992).7

Community-based nonprofit organizations are viewed to be especially beneficial for innovation and localized, community-based development. They provide an opportunity for small-scale individualized development, which can be essential for some communities (Rubin, 1993). Because of their ability to address both the technicalities of physical development and the sensitivities of human/social service, community-based organizations are viewed as essential to neighborhood revitalization and a necessary player in the housing arena (Walker, 1993). Some researchers maintain that this assertion extends to FBOs as well, signaling a greater governmental desire and willingness to partner with faith-based and community-based organizations and to entrust them with the delivery of key public services. This trust may be attributed, in part, to the belief that nonprofit organizations in general and faith-based and

105

community-based organizations in particular embody a unique set of resources and strengths that can facilitate their work in the housing field (Vidal, 1995). For instance, FBOs (1) tend to be trusted by their communities, (2) can access both human and financial capital, (3) are cultural anchors within the community, (4) are capable of providing and creating leadership within the community, (5) have the capacity for relatively quick implementation, (6) are able to customize programs and services to address local conditions, and (7) can reach individuals and communities that might not otherwise be served (Bratt et al., 1994; Chaves & Tsitsos, 2001; Rubin, 1993).

Organizational Considerations

The nonprofit housing sector continues to be diverse on a number of dimensions, including organizational and program characteristics, units produced, and level and sources of funding (National Congress for Community Economic Development [NCEED], 1999, 2005; Walker, 1993). The following discussion considers some of these similarities and variations.

The extant research suggests a number of factors that may affect the extent and nature of an organization’s housing and other community service activities, including organization age, size, financial status, and location (Chang, Williams, Griffith, & Young, 1994; Chaves & Tsitsos, 2001; Devita, Platnick, & Twombly, 1999; Hall, 1992; La Barbera, 1992; Lincoln & Mamiya, 1990; Reese & Shields, 2000).8 For example, older, larger, and more financially stable FBOs, which are often located in urban centers, have been more likely than other kinds of organizations to engage in service delivery. But does this apply to engagement in the housing arena, or are there other more pressing factors?

Organizational Staffing

A variety of nonprofit organizations function in the housing industry. For example, community development corporations (CDCs), both faith-based and secular, are central to housing provision. They range in staff size from as few as 10 to well over 1,100. This variability can make it difficult to compare capacity across organizations (National Alliance of Community Economic Development Associations [NACEDA], 2007). In 1990, the median staff size of CDCs nationally was seven paid employees (Walker, 1993). A 2005 national survey of CDCs found similar patterns, with the median size for these organizations of seven full-time staff, three part-time staff, and five volunteers (NCCED, 2005). Similar staff sizes have been observed for other types of faith-based and secular human service agencies (Clerkin & Grønbjerg, 2007). However, research indicates that FBOs and secular organizations involved in housing services for the homeless in Houston, Texas, had a mean of 27 and 39.9 paid staff, respectively. Furthermore,

106

FBOs were twice as likely to rely on volunteers rather than paid staff, whereas secular organizations had a 0.89:1 ratio of volunteers to paid staff (Ebaugh, Pipes, Chafetz, & Daniels, 2003). In general, research suggests that FBOs rely more heavily on volunteers and have a lower proportion of their budget covered through government funding than do their secular counterparts (Chaves, 2002; Cnaan, 1997; Ebaugh et al., 2003; Ebaugh, Pipes, & Chafetz, 2006; Hodgkinson & Weitzman, 1993; Monsma, 2002; Printz Platnick, 1998; Twombly, 2002). Some of the research shows a connection between paid staffing, use of volunteers, and percentage of government funding; nonprofit organizations that receive more government funding tend to rely more on paid staff for critical organizational functions and less on volunteers (Salamon, 1987a, 1987b; Smith & Lipsky, 1993; VanTil, 1988). The potential variability in organizational size, staffing, and funding merits additional attention.

Funding

Nonprofit funding has received increasing attention in the research literature (e.g., Crittenden, 2000; Ebaugh, Chafetz, & Pipes, 2005a, 2005b; Grønbjerg, 1993; Salamon, 1987a, 1998; VanTil, 1988). Faith-based and secular organizations rely on a variety of funding sources, including the government, banks, foundations, individuals, and a range of other secular agencies and religious organizations (Walker, 1993). While FBOs tend to rely heavily on religious funding sources, secular orgnizations tend to rely more heavily on secular sources and the government for funding (Burt et al., 1999; Ebaugh et al., 2003). This may be due to a fear on the part of FBOs that receiving government funding will jeopardize or diminish religious missions (Chaves, 1999b; DiMaggio & Powell, 1991). However, evidence is mixed as to whether organizations actually alter their priorities, goals, and/or activities to satisfy funders (Ebaugh et al., 2005b Garland, 1992; Monsma, 1996; Smith & Lipsky, 1993; Smith & Sosin, 2001).9

The Housing Arena

Nonprofit organizations, whether faith-based or secular, can play a variety of roles in housing, which are not mutually exclusive: (1) providing actual housing units, (2) delivering wraparound and auxiliary housing services, (3) building political support for affordable housing,10 and (4) undertaking a variety of broader community development activities (Briggs, 2004). The variety of ways in which nonprofit organizations can be involved in the production of housing units is summarized in Table A-1 (see Appendix A).

Housing Production

CDCs play a particularly vital role in the local landscape from housing construction and rehabilitation to real estate, industrial and small business development, to creating job opportunities

107

and a range of other supportive social services. Since the 1960s, CDCs have proliferated, including those established by and within the faith-based community. By the early 1990s, 2% of CDCs engaged in housing production were producing 25% of the sector’s housing units. Nearly 50% of the CDCs accounted for less than 8% of units placed in production by the sector. Tables A-2 and A-3 (see Appendix A) highlight the housing units produced by CDCs over time as well as the growth in the number of CDCs and the proportion of FBOs engaging in the housing arena. Although once concentrated in the urban centers of northeastern and central states, by the early twenty-first century, CDCs were equally dispersed across the United States, with representation in rural areas and small cities as well as the traditional urban core (NACEDA, 2007; Walker, 1993). This geographic dispersion suggests an increased presence and opportunity for faith-based and secular involvement in housing services.

Capacity Building and Technical Assistance

Faith-based and secular nonprofit organizations have noted a number of obstacles to their involvement in the housing arena, including financing and organizational capacity issues, such as training and technical assistance and general operating support (Vidal, 1992; Walker, 1993). Scholars and practitioners agree that building organizational capacity and providing sufficient and timely technical assistance to nonprofit organizations is vital to their continued success and ability to provide services in a systematic and sustained manner. In the housing arena, some CDCs have been reluctant to acknowledge capacity problems out of a fear that such an admission would negatively impact funding (Bratt & Rohe, 2005). Fortunately, technical assistance can come from a variety of other sources, including universities, consulting firms, or other private organizations. There has been a concerted effort on the part of all of these entities to increase technical assistance that will enhance the capacity of nonprofit organizations involved in the housing arena. The extent to which organizational capacity can be built will depend in large part on the willingness of the organization to seek and obtain necessary assistance.

Partnerships and Collaboration

There are a variety of reasons why nonprofit organizations enter into partnerships and collaborations to provide housing services. For example, resource constraints may encourage neighborhood-based and/or small organizations to cooperate with other organizations to carry out programmatic goals (Reese & Shields, 2000). Many organizations, regardless of size, find that they need to partner with various funders to accomplish tasks and goals in the housing arena.11

108

Service providers can take distinct, categorical approaches to housing or more integrative, holistic approaches. Increasingly, nonprofit organizations are opting for the latter. Collaboration between government, nonprofit, and for-profit housing providers occurs at each level of government. The importance of collaboration and networks is even stronger for FBOs and other nonprofit organizations that are challenged in their capacity to provide housing units and related support services. For example, these organizations can increase their effectiveness and impact if they collaborate to (1) build an effective and representative board, (2) retain program staff in the face of uneven funding levels, (3) share expertise and other resources, (4) avoid duplication, (5) create linkages to disenfranchised constituencies in their communities, and (6) forge political alliances (Mitchell, Florin, & Stevenson, 2002, pp. 627–628).

Even with the potential benefits, nonprofit organizations are likely to find some hindrances to collaboration. For example, the mission, goals, and intended beneficiaries of participating organizations may be sufficiently distinct to cause some hesitation or unwillingness to work together on the part of potential collaborators. Funding sources and associated conditions may also make it difficult for some organizations to work together even when there is a sincere desire to do so (Guo, 2007; Mitchell et al., 2002). Organizations that already have a working knowledge of each other and a sense of trust are more likely to work together. This suggests that organizations may have some opportunity costs in developing necessary foundations to collaborate on housing and other projects.

RESEARCH AIMS

At its most general level, this analysis seeks to discover whether there are systemic differences between FBOs and secular nonprofit organizations other than obvious faith orientation, and whether such differences impact the production and distribution of low-income housing. This question is at the heart of the public debate as to whether efforts to increase FBO engagement in the public sector are desirable. This preliminary analysis is organized around four specific questions:

1. How do FBOs compare with secular nonprofit housing organizations on general organizational characteristics? This analysis tests the possibility that differences between FBOs and secular organizations are actually independent of faith orientation and can be better explained by traditional organizational variables, such as size, age, and budget.

2. How do reported housing production measures generated by FBOs compare with secular nonprofit housing organizations? Output is measured on a variety of housing-specific indicators as well as broader measures of organizational activity.

109

3. How do FBOs and secular nonprofit organizations perceive constraints on expanding organization capacity? Organizational-level perceptions of constraints provide important guides to the design of appropriate policy to further engage nonprofit organizations (both FBO and secular) in the low-income housing sector. This analysis also provides insight into organizational priorities of FBOs and secular nonprofit organizations and determines whether there are systematic variations in these priorities.

4. To what extent do FBOs and secular nonprofit organizations rely on similar networks of partners and collaborators? This analysis compares working partners and collaborators of FBOs and secular nonprofit organizations, and provides preliminary evidence as to whether different organizational types mobilize different resources and collaborations into the overall public effort to provide low-income housing.

DATA AND METHODS

The data reported in this paper were obtained from the Community Economic Development Association of Michigan (CEDAM), a nonprofit organization which is serving as an organizational partner for this research.12 CEDAM provided raw data from a survey conducted via telephone and mail from October 2003 to April 2004. The survey population comprised CDCs in Michigan that were identified through multiple databases, including those of CEDAM and members of the Michigan Training and Technical Assistance Collaborative (MTTAC). A total of 176 surveys were distributed, with a 51% response rate.13 The regional distribution of respondents matched that of the membership base throughout the state. Appendix B provides additional information on the survey population and methods.

The survey aimed to collect information on the impact of CDCs in Michigan’s neighborhoods and communities. Although it is possible that there are CEDAM members that provide no housing or housing-related services, only 23 of the survey respondents produced no housing units and only 13 respondents provided no housing-related services. The survey collected baseline data on housing production, commercial development, open-space and/or green-space development, and community facilities development. Data were also collected on the services associated with housing, community building, and economic development provided by CDCs as well as technical assistance, capacity building needs, and general demographics. Respondents included both faith-based and secular CDCs;14 thus, it was possible to isolate and compare the housing and community development activities of faith-based and secular CDCs as well as the service activities associated with those developments.

The use of CEDAM as a partner organization inherently limits the empirical focus of the initial portion of this research to the state of Michigan, an issue worthy of some discussion because

110

admittedly it reduces the generalizability of the findings. However, there are several advantages to a single-state approach. First, examination of changes in faith-based service provision within a state controls for state legislation in housing, community development, and public funding conventions. A single-state focus allows for in-depth understanding of the methods and outcomes of faith-based service provision. The community focus within a state captures the essence of service delivery activity in the housing arena and provides the type of in-depth analysis essential for understanding the institutional and organizational changes in power relationships resulting from innovations in service frameworks.

Second, Michigan represents several ideal factors that permit generalizing results. While often characterized as a classic Rust Belt state, there are many areas of Michigan that remain rural and/or where the primary source of income is tied to tourism and agriculture. Michigan, the eighth most populous state, also has a high percentage of minorities and ethnic groups concentrated in distinct geographical regions. There are areas of the state that are distinctly more liberal, whereas other areas are conservative in both political and lifestyle preferences. Many blue-collar communities in Michigan experienced in-migration from the South only a generation or two ago and remain distinctly working class, whereas other areas in the state have seen a suburban explosion of middle-class and upper-middle–class wealth. Thus, Michigan embodies a microcosm that is sufficiently heterogeneous in client groups and service providers to provide a good sense of changing modes of public service provision.

FINDINGS

There is a good deal of speculation in both the research and popular literature that faith-based and secular nonprofit organizations differ on many basic organizational characteristics. In the following analysis, four distinct dimensions of these organizations are compared.15 The first focuses on structural characteristics, such as size and organization age. A second set of indicators examines organizational activity and focus. The third dimension measures organizational perspectives on the environment in which it must operate. The fourth set of indicators presents evidence on actual levels of housing produced by each institutional type.16

Table A-4 (see Appendix A) compares faith-based and secular nonprofit housing organizations by years in operation, total employees, and annual budget. Although this table shows a great deal of variation in these basic structural indicators, secular status explains little of the observed variation. However, the central tendency data in this table are somewhat misleading in that they include a large faith-based nonprofit organization, which results in a set of average scores that give the impression that the “average” FBO is significantly larger than it actually is. Within the

111

table, the values presented in parentheses are the recalculated means if the single large organization is dropped from analysis. The recalculated means suggest that FBOs are somewhat smaller and have fewer employees than their secular counterparts. However, these differences are of smaller magnitude than differences within each category of organization, and certainly do not rise to the level of statistical significance. Interestingly, in this study, secular organizations relied on volunteers more than they relied on paid staff, as well as more than FBOs, which runs counter to findings in previous studies.

Organizations were asked to identify which of 43 specific activities were areas in which they were active. These activities included housing (16 examples), business development (9 examples), social services (14 examples), and “other” (4 examples). Figure A-1 (see Appendix A) highlights examples of the range of housing-related activities in which these nonprofit organizations engage. Interestingly, the data indicate that faith-based and secular nonprofit organizations do not provide substantially different kinds of housing services, but rather engage in very similar activities. For a more comprehensive view, Table A-5 (see Appendix A) compares FBOs and secular organizations with respect to the number of functionally distinct activities they report; once again, the findings show little difference between faith-based and secular nonprofit housing organizations.

A closer analysis of specific activities, however, shows some differences between FBOs and secular organizations; namely, secular organizations exhibit a greater propensity to engage in advocacy activity, micro-business development, and business district development. FBOs are more likely to engage in leadership training, arts programs, and business operation.17 Perhaps the most important comparison to be made between faith-based and secular housing nonprofit organizations is the actual production of services. The data reported in Table A-6 (see Appendix A) present small but very interesting differences between FBOs and secular organizations. In all cases, FBOs show somewhat greater levels of housing output (i.e., projects, units, and future units). This difference is greatest for future projects. The implication is that if these projections turn out to be accurate, then the gap between faith-based and secular housing unit production will increase in the future.

Measuring the Capacity of Faith-based and Secular Nonprofit Housing Organizations

Contrary to what the literature suggests, organizations are quite willing to identify barriers or obstacles to their involvement in the housing arena. Among 90 organizations in the study sample, 79 (88%) reported at least one barrier to expanding the scope and capacity of their organization. Also, 50 organizations (56%) pointed to specific barriers within the local planning

112

process, and 32 (36%) identified barriers to obtaining needed training and technical assistance. Figure A-2 (see Appendix A) demonstrates that once again there is a strong agreement between faith-based and secular nonprofit providers as to major impediments to growth and expanded organizational capacity. While it is not surprising that there is significant consensus that the major impediment faced is a lack of operating funds, there is also a remarkable coherence across organizational types in perception of additional problems. FBOs and secular organizations perceive the same barriers to growth in housing production: operating funds, administrative capacity, land acquisition, and resources to fund projects.

One indirect indicator of organizational capacity is the range of deficiencies identified by organizations themselves. As noted above, the CEDAM survey asked nonprofit housing organizations to identify needed areas of technical assistance and training. These questions included, of course, housing concerns, but they also mentioned possible topics in community planning, commercial development, and commercial strip development. Faith-based and secular nonprofit housing organizations indicated similar preference to 19 of 20 possible areas of technical assistance. The single exception to this pattern for housing-related services was technical assistance in the area of tax credits. Here, FBOs expressed a need for more technical assistance than their secular counterparts. FBOs also expressed a need for more technical assistance for neighborhood planning and market studies in the area of community planning, which could have implications for housing activity. Figure A-3 (see Appendix A) highlights some of the key technical assistance needs identified by the organizations. There were no statistically significant differences with respect to 13 other categories of proposed training. There were, however, interesting differences in how organizations perceived barriers to obtaining needed technical support. Lack of knowledge, lack of time, and staff capacity were all cited as barriers to obtaining technical assistance.18 FBOs are much more likely to identify funding as a major barrier to obtaining technical assistance. Interestingly, FBOs are much less likely to identify staff capacity as a problem, an issue worth pursuing in future research.

A similar pattern emerges with respect to perceived barriers to organization and project planning, providing further evidence that FBOs and secular organizations see different barriers to success. FBOs are more sensitive to perceived political and bureaucratic barriers than are secular organizations. FBOs also see more problems with local regulations, such as city ordinances, and FBOs are somewhat more concerned with funding issues.19

113

Organizational Networks

A key issue surrounding efforts to engage FBOs in the production and distribution of publicly financed goods and services is the long-term impact such engagement might have on the potential for cooperative arrangements among low-income housing providers. This is, of course, an extraordinarily difficult question to examine empirically. Figure A-4 (see Appendix A) summarizes the reported partners by faith-based and secular nonprofit housing organizations. Again, there are some interesting differences between FBOs and secular organizations. Among FBOs, 10% reported banks as partners compared with 4% of secular organizations. In addition, FBOs are more likely to partner with “other” nontraditional partners (24% versus 8%). In contrast, secular organizations are more likely to partner with public agencies. This pattern holds for federal, state, county, and city agencies. The difference between secular organizations and FBOs is particularly clear for state agencies (16% versus 7%), county agencies (8% versus 0%), and city agencies (16% versus 10%).20 There is no evidence as to whether variation in project partners represents organizational preferences or is somehow imposed by external constraints.

Figure A-4 also reports responses to questions about “problems” in obtaining financial support for projects. If agencies are seen as sources of problems, then a failure to partner with them likely has more to do with institutional constraints than internal preferences. Secular and faith-based housing providers have similar views of private foundations, corporations, and federal agencies. Approximately 20% to 22% of all nonprofit housing organizations see such organizations as sources of problems in financing projects. However, there seems to be no systematic difference between secular organizations and FBOs in expressing such concerns. Dissatisfaction with state agencies is a bit lower, with 18% of secular organizations and 14% of FBOs expressing concern. Major differences appear with respect to city and county agencies, where FBOs are much more likely to perceive city and county agencies as serious barriers to potential project financing. It is unclear whether these perceptions are the result of prior negative experiences or a general mistrust of government agencies. Future research and case study interviews will address this issue.

Summary of Findings • Overall, this analysis revealed very little difference between FBOs and secular

organizations in what they do in the housing arena and how they do it.

• Faith-based housing providers anticipate more housing production in the future, based on projects currently planned. Whether this reflects optimism or is linked to actual commitments is difficult to discern at this point.

• Monetary resources are a greater barrier to FBOs than to secular organizations in housing production and service provision.

114

• Faith-based housing providers need more technical assistance on tax credit housing, neighborhood planning, and market studies than do secular housing providers.

• Faith-based housing providers are much less likely than secular housing providers to identify staff capacity as a problem.

• FBOs are more sensitive to perceived political and bureaucratic barriers in the housing arena than are secular organizations. FBOs also see more problems with local regulations.

• FBOs are more likely to have banks as partners and are more likely to partner with other nontraditional partners. Secular organizations are more likely to partner with public agencies.

• Faith-based housing providers are much more likely than secular housing providers to perceive city and county agencies as serious barriers to potential project financing.

Even with these noted distinctions, considerable similarities remain between FBOs and secular organizations in the provision of housing services.

IMPLICATIONS FOR RESEARCH, POLICY, AND PRACTICE

While the public debate on the appropriate role of FBOs has been extensive and frequently heated, resolving key issues has been difficult because of a lack of reliable data on what FBOs do and how they do it. Indeed, the information base for secular nonprofit organizations is hardly any better. While some of the limits to the generalizability of data presented here have been noted, the analysis does offer two intriguing possibilities. First, much of the debate about the appropriateness of engaging FBOs relative to secular organizations in service delivery may overestimate differences between them. The two types of organizations are in fact quite similar in terms of structure, resources, partners, and perceived problems. These findings support the notion that FBOs may make good partners with government agencies, as FBOs appear to be at least as productive as their secular counterparts, and if future projections hold, these organizations could generate relatively more housing than secular organizations in the future. This finding is particularly salient given that faith-based nonprofit organizations are often smaller and underfunded relative to their secular counterparts. Indeed, it suggests that if observed disconnects between FBOs and potential partners (particularly local government) can be repaired, there is a high potential for increased housing service production by FBOs.

This analysis provides a framework for improved understanding of differences in service activities, and capacities, of faith-based and secular housing providers. It also provides a foundation to develop a better understanding of networks of U.S. housing-services providers. It is important to continue to highlight the unique capacities and roles of FBOs in housing service

115

provision. This and related research is important because an improved understanding of FBO activities and capacities can help identify capacity deficits that can be addressed through fundiand training. Similarly, understanding service-provision networks will enhance the ability of federal, state, and local policy makers to target assistance and incorporate into policy the realitythat services are provided by different types of actors. As such, it continues to be important to identify areas of potential cooperation between public, faith-based, and secular service provid

Research

ng

ers.

research has found that faith-based and secular housing providers are similar in their

ve poorer, more t

• the quality and nature of services provided by

• y networks among housing providers

ers

The researc to enhance understanding of

ionships with governmental actors

• examine why FBOs are more likely to see local ordinances and

d s

Overall, theactivities and approaches, with a few interesting differences. There are, however, a number of research questions not addressed by the data at hand, and attention to these may result in more nuanced findings that may indeed show important differences between faith-based and secular service providers. Several of these missing questions include the following:

• Examination of clients served—it is very possible that FBOs serchallenging, or more isolated clients or neighborhoods, thus filling a policy gap lefby secular and government providers.

Assessment of client satisfaction withdifferent types of organizations is still necessary to determine if there are identifiable differences in the quality of services provided.

More extensive and detailed analysis of policmay show different cooperative approaches between faith-based and secular providers that have implications for service capacity. In addition, such analysis is needed to assess whether fostering networks among faith-based and secular providmight expand capacity for both types of organizations.21

h also points to other questions that need to be addressedthe interface between government and faith-based service providers. Such understanding will form the basis for improved public policy related to both funding and implementing housing provided by FBOs. These questions include the following:

• Secular organizations appear to have better relatand seem more likely to have government funding. Is this the result of choice (e.g., FBOs are less comfortable with public funding because of the strings attached or the fear of diluting religious mission) or the result of systematic or unintended bias in the funding system? Research should explore the reasons FBOs are less likely to have public funding.

Research shouldzoning as barriers to housing provision. Are FBOs less skilled at addressing such issues? Is there something systematically different about the location of faith-baseprojects that leads to greater difficulty getting necessary permits? Are the problem

116

with local ordinances among FBOs related more to housing projects than services? Abetter understanding of the barriers or perceived barriers would allow public actors tbegin to address these concerns.

Practice

o

Policy and

tant research questions remain, the findings here have several implications

lations as more problematic. To increase their role in housing

hnical assistance in several market-related aspects of housing

important and relevant policy finding of this research is that faith-based

Although many imporfor policy and practice relating to the interface between faith-based and public provision of housing services. The first issue relates to the question of public financing of faith-based service providers. Despite federal and state initiatives embodied in Charitable Choice and other programs, faith-based housing providers are less likely to use government funding (and concomitantly are more likely to partner with banks) and indicate greater concerns about and problems with interacting with government and governmental agencies. Regardless of whether these concerns are real or perceived, there appears to be a need for some sort of public action to smooth out these relationships.

FBOs see local politics and reguprovision, state and local actors may need to do more to alleviate some of these barriers or at least the perceptions of barriers. Such activities could include making FBOs more aware of funding opportunities and regulations, creating FBO ombudspersons or offices within state and local government agencies to facilitate funding, working with government regulations or red tape, zoning, and building permits, as well as government facilitation of partnerships between faith-based and secular providers.

FBOs also appear to need more tecprovision: market studies, specific state housing program regulations, and local planning and zoning. This could be provided through workshops, but could also be alleviated by a cooperative or mentoring program where secular housing providers work with their faith-based counterparts on these issues.

Perhaps the mostproviders expect significantly more housing output than secular organizations in the future. Furthermore, FBOs are much less likely to identify staff capacity as a barrier and rely on and have significantly more volunteers. This suggests that at the least there is a still growing capacity in the faith-based sector for housing and perhaps other social services as well. If some of the funding and governmental relations barriers could be addressed, this capacity would be better realized.

117

ACKNOWLEDGMENTS

This research was funded by a grant from Michigan State University’s Community Vitality Program. We appreciate the assistance of graduate assistants, Davia Cox, Jessica Faist, and Rebecca Bromley in the early stages of this research.

NOTES

1. This analysis is part of a larger interstate project that explores the role of faith-based service provision in urban areas given changing political, market, social, cultural, organizational and institutional forces; assesses the extent of convergence in how and what FBOs, secular nonprofit organizations, and governmental institutions do in meeting the needs of urban populations; and examines the possible long-term impact of faith-based service provision on urban institutional arrangements.

2. Many of the organizations included in the study are community development corporations (CDCs).

3. Attention is given to FBOs in this study because they are involved in the provision of a number of social services, including housing, education, childcare, health, and economic development (Roozen, McKinney, & Carroll, 1984; Wineburg, 1996). Housing is a particularly important area of FBO involvement in social services provision for a number of reasons, including the reality that it is a service where FBOs have a long history of involvement. FBOs also use their experiences with housing services as a springboard to other community and economic development activities, including job and entrepreneurial training as well as business incubation and consultation (Cisneros, 1996; Heim, 1995; Hodgkinson & Weitzman, 1993; La Barbera, 1992; Lincoln & Mamiya, 1990; Reese & Shields, 1999, 2000).

4. Governmental and societal acceptance of nonprofit organizations in service delivery and other typical government roles is the result of two competing forces: the increasingly negative view of government and its capacity, coupled with the generally positive view of the ability of nonprofit organizations. Streams of this argument are evident in U.S. social policy passed at the federal, state, and local levels since the 1980s and continue to emerge in current policy discussions. While the case could be made for a discussion of faith-based and community-based organizations; the focus of this research is specifically the role of FBOs in housing provision.

5. See Figure A-5 (Appendix A) for highlights of select historical events that impacted the role of nonprofit organizations in the housing arena.

6. This included Section 221(d)(3) and 236 below-market-interest-rate programs: the Office of Economic Opportunity and the Model Cities Program, which was supportive of housing development corporations; the 1966 Special Impact Amendment to the Economic Opportunity Act, which officially led to the creation of CDCs; and Title IV of the 1974 Community Services Act, which authorized considerable funding for CDCs (Keyes, 1971; Mayer, 1984). Several federal demonstration programs that were eventually folded into the Community Development Block Grant program have provided considerable opportunities for CDCs in the housing arena.

118

7. Although there have been numerous opportunities for nonprofit organization involvement in housing, increasing pressure for community voice, combined with the privatization movement created additional pathways for nonprofit organization involvement in the housing arena (Boyte & Riessman, 1986; Hodgkinson & Lyman, 1989). From one perspective, one might argue that the introduction of Charitable Choice legislation and President Bush’s faith-based and community initiatives were simply two more federal efforts to enlarge opportunities for nonprofit organizations. The caveat, of course, is that different types of nonprofit organizations were potentially gaining access. Although religious organizations had been subsumed broadly within the nonprofit realm, these latter federal initiatives specifically reach out to both faith-based and community-based organizations.

8. The research also indicates that with respect to FBOs, denomination, theology, pastoral characteristics, pastoral initiative, extent of political activities, and being located in a census tract with low educational attainment affects community service activities.

9. Where the impact may be noticeable is with regard to networking activity. Guo (2007) found that nonprofit organizations that relied on government funding were less likely to develop governing boards with strong community representation. This, in turn, might impact the ability of that organization to adequately network with other individuals and organizations in the community.

10. Building political support includes activities like political advocacy, grassroots organizing, and building coalitions with other organizations.

11. Organizations involved in housing production and the delivery of housing services work with an array of financing institutions, government grant and loan programs, foundation grants, national nonprofit intermediaries (e.g., Local Initiatives Support Corporation [LISC], Enterprise Community Partners, NeighborWorks America) as well as corporate and individual giving.

12. CEDAM is a voluntary association of CDCs, individuals, and other organizations focused on promoting and expanding community-based housing and economic development across the state of Michigan. CEDAM activities include building the capacity of CDCs and other community development partners; generating public and private resources for CDCs; providing a forum for member input in public policy debates; and bringing local, regional, and national organizations together to foster community-based economic development (CEDAM, 2004).

13. This is an adequate response rate and certainly within the range of response rates for other mail surveys administered to nonprofit organizations since the late 1990s (Babbie, 1990; Hager, Wilson, Pollak, & Rooney, 2003).

14. It should be noted that the survey did not contain a question indicating whether an organization was faith-based. Responding organizations were contacted via telephone and e-mail to identify the extent and nature of their religious or secular status. See Appendix B.

15. Table C-1 (see Appendix C) provides a comparison of faith-based and secular organizations on a number of dimensions.

16. Factor analysis was run to determine if there were any underlying constructs and revealed that all the activities loaded on a single factor.

119

17. See Tables C2 through C5 (Appendix C) for additional details.

18. See Tables C7 thorough C9 (Appendix C) for additional details.

19. See Table C-10 (Appendix C) for details. The implications of these barriers, real and/or perceived, are addressed elsewhere in this paper.

20. See Tables C11 and C12 (Appendix C).

21. It is important to note that these issues were excluded from the baseline survey conducted by CEDAM, but not from the researchers overall research agenda or subsequent data collection.

APPENDIX A: TABLES AND FIGURES

Table A-1. Select Dimensions of Housing Unit Production, by Nonprofit Organizations

Dimension Example

Legal/Financial Arrangement

• Owner-developer

• General partner

• Limited partner

Level of Development

• New unit creation

• Unit rehabilitation (substantial, moderate, light)

• Repair

Type of Unit

• Owner-occupied

• Single-family rental property (1–4 units)

• Multifamily rental property (medium- or large-scale building)

• Cooperative

Institutional Relationships

•(Partnerships/Cooperation)

• Public sector

• Financial institutions

Nonprofit organizations

Extent of Involvement nt

• Ad hoc (project-by-project)

• Ongoing program manageme

Source: Walker (1993)

120

Table A-2. Percentage of Community Development Corporations Producing Housing Units between 1988 and 1990

Number of Housing Units Percent

Up to 10 48.7

11–25 26.5

26–50 14.5

51–100 5.7

101–200 2.7

More than 200 1.7

Table A-3. Community Development Corporation Proliferation and Housing Production, 1998 and 2005

1998 2005

Number of CDCs 3,600 4,600

Percent faith-based CDCs 14% 25 %

Housing units produced by CDCs 650,000a 1.2 million a An estimated 109,000 of those housing units had been produced by FBOs, accounting for approximately 17% of

all housing units produced by nonprofit community organizations at that time. Using similar data from 1991, Walker (1993) estimated that by 1990, CDCs (no distinction between religious and secular) had accounted for 15% of all housing units produced. For the 1995 data, Habitat for Humanity accounted for another 30,000 units, resulting in over 31% of housing units produced that can be attributed to the faith-based community.

Source: (Fannie Mae Foundation, 2001; NCCED, 1999, 2005).

Table A-4. Structural Characteristics of Secular and Faith-Based Housing Providers

Organization Type N Mean Std. Deviation

t-Test Significance

Secular 48 21 12 Number of Years in Operation Faith-based 27 19 8

.45

Secular 46 15 40 Total Employees Faith-based 27 25

(6) 95 .53

Secular 41 .0140 .07 Ratio of Volunteers to Employed Workers Faith-based 23 .0017 .01

.01***

Secular 42 $2,323,652 $3,434,243 Annual Budget Faith-based 26 $3,728,840

($1,157,993) $13,217,195 .51

*** Significant at the 0.01 level ( ) recalculated values after dropping one very large faith-based nonprofit.

121

Figure A-1. Housing Activities of CEDAM Membersa

a This figure includes home ownership counseling, which the CEDAM survey characterized as a social service and

8 of the 16 housing development activities. The 8 housing development activities included on the survey that are not identified in this figure include HOPE VI (2.2%), land trusts (7.8%), SRO housing (6.7%), asbestos removal (5.6%), lead paint (25.6%), elderly housing (22%), AIDS housing (13.3%), and disability housing (17.8%).

Table A-5. Activities of Secular and Faith-Based Housing Nonprofitsa

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 51 7.3 4.0 .56 Total Number of Reported Activities Faith-based 29 6.9 4.1 .76

.717

Secular 51 3.4 3.1 .43 Total Number of Reported Housing Activities Faith-based 29 3.3 2.8 .52

.631

Secular 51 3.4 3.1 .43 Total Number of Reported Nonhousing Activities Faith-based 29 3.3 2.8 .52

.912

a The differences were measured by computing a t-test between the mean number of organizations within each type reporting activity in that area.

122

Table A-6. Housing Outputa from Secular and Faith-Based Housing Nonprofits

Organization Type N Mean Std. Deviation

t-Test Significance

Secular 41 2.2 1.4 Total Number of Existing Housing Projects Faith-based 19 2.8 1.2

.11

Secular 51 57 89 Total Number of Units in Existing Projects Faith-based 29 63 84

.77

Secular 51 34 62 Total Number of Units in Future (i.e., planned) Projects Faith-based 29 63 87

.09*

* Significant at the 0.10 level. Significance levels at 0.10 are reported due to the relatively small number of respondents. It is possible that the study is underpowered to detect true relationships, if any exist. A larger number of respondents may have revealed greater differences.

a Output represents housing projects, units in existing projects, and projected units in future projects.

Figure A-2. Impediments to Growth and Expanded Organizational Capacitya

a See Table C-6 (Appendix C) for additional details.

123

Figure A-4. Partners and Sources of Financial Problems

124

Figure A-3. Technical Assistance Needs

Figure A-5. Historical Overview of Select Events and Activities Influencing Nonprofit Organization Involvement in the Housing Arenaa

Late 1800s– Early 1900s 1959

1962– 1967 1966 1968 1974

Mid-1970s– 1980s 1980s 1988 1990s 1996 2001

Settlement H

ouses and Cooperatives for fam

ilies

Federal Section 202 program (affordable housing, disabled, elderly)

Ford Foundation Gray A

reas pilot program

Special impact am

endment to the 1 964 Econom

ic Opportunity A

ct (creation of com

munity developm

ent corporations [CD

Cs])

Dem

onstration Cities and M

etropolitan Developm

ent Act (M

odel C

ities Program)

Housing and U

rban Developm

ent Act (Section 236 program

)

Com

munity Services A

ct—Title IV

H

ousi ng Act (Section 8)

Widespread interest in privatization; p articularly in housing, w

ith increasing focus on nonprofits

Federal housing assistance decreased; N

umber of com

munity based housing providers increased

Bowen v. K

endrick—U

. S. Supreme C

ourt decision affirming faith-

based organization involvement in service delivery

Renew

ed interest in nonprofit service provision, growth in C

DC

s and ho using production by C

DC

s

Charitable C

hoice Com

ponent of Personal Responsibility &

Work

Opport unity R

econciliation Act

Executive Orders establishing Faith-based and C

omm

unity Initiative (FB

CI) and centers to coordinate FB

CI activity

a This overview is intended to be illustrative and not exhaustive.

125

APPENDIX B: ADDITIONAL INFORMATION ON DATA AND METHODS

DATA

The survey used for this analysis was based on the population of CEDAM and Michigan Training & Technical Assistance Collaborative (MITTAC) members; thus, it is a census of all organizations of this type. This membership list and hence the survey population included three fourths of the community-based housing providers in Michigan. Thus, while it did not represent the population of low-income housing providers in the state, it was sent to the complete membership of CEDAM/MITTAC, the only extant list of low-income housing providers. It is important to note that the representativeness of this sample of organizations that provided housing services in Michigan was actually higher because community-based housing providers were the organizations that actually joined CEDAM. Because this membership list was as close to a census as one could reasonably expect, no sampling was required. CEDAM has six regions organizing their membership. The regional location of survey respondents closely matched overall member location, indeed the ranking of regional representation was identical (see Table B-1). Given the representativeness of respondents no weighting was necessary.

Table B-1. Regional Distribution of Survey Respondents and CEDAM/MITTAC Membership

Michigan Region Respondents Population

Upper Peninsula 6% 3%

Northern Lower/Thumb 6% 5%

Southwest 15% 15%

Central 10% 13%

Southeast 23% 20%

Detroit 40% 44%

N 90 176

CEDAM’s membership comprises Community Development Corporations (CDCs), Community Housing Development Organizations (CHDOs), financial institutions, city governments, state agencies, consultants, national intermediaries, community action agencies, and various Habitat for Humanity affiliates around the state. Members include the following:

126

• Community development corporations

• Other community-based nonprofit organizations

• Community action agencies

• Municipalities

• Financial institutions

• Training and technical assistance providers

• Consultants

• Higher education institutions

• Private developers

• Law firms

• Concerned individual community members

MITTAC comprises organizations and agencies providing training and technical assistance to Michigan nonprofit organizations and local units of government that are engaged in sponsoring, producing, and/or operating affordable housing and other community development activities.

The raw survey data were provided to the research team by CEDAM as an Excel spreadsheet containing quantitative and qualitative data. The complete CEDAM survey is available at www.gusp.msu.edu. The research team cleaned the data, developed a corresponding codebook, and prepared data files suitable for quantitative analysis using the software SPSS and qualitative analysis by the software NVivo7. A cursory review of the data indicated that respondents likely included both faith-based and secular CDCs; however, there was no coding to differentiate between faith-based and secular service providers in the original data set. The research team contacted each organization that had responded to the CEDAM survey to determine whether it was faith-based or secular and this variable was added to the database. Common sense suggests that the faith orientation of an organization cannot be determined by name alone. Although it may be less apparent, a simple examination of organizational mission statements is also fallible because an organization may not be explicit about its religious intent and purpose. Empirical findings support this as well (Ebaugh, et al., 2003). This research uses self-identification measures to determine whether an organization is faith-based or secular.

Among the 90 organizations, 29 identified themselves in the follow-up phone call as faith-based, 51 as secular, and 11 organizations could not be reached to verify their status. Based on this organizational self-identification, it was possible to isolate and compare the housing and community development activities of faith-based and secular CDCs as well as the service activities associated with those developments.

127

CONTEXT

This research is the initial part of an ongoing project focusing on the impact of faith-based service provision and urban governance. Although the larger research employs surveys, face-to-face interviews, and case studies, only baseline survey data provided by CEDAM are presented here (an original survey designed by the research team and case studies are still ongoing). The research is essential to the study of faith-based service provision for two reasons. First, there is a dearth of extant quantitative research on the topic. While the national study by Chaves (1999a) provides critical information on faith-based community development activities, the questions were not designed to obtain the detailed service information necessary to understand the changing nature and extent of faith-based service and effects of government funding. Second, many existing case studies provide limited and nongeneralizable snapshots of the activities of just a few, often atypically large and well-resourced, FBOs. The overall research design builds on existing databases to provide generalizable data, amenable to statistical analysis, on FBO service providers. It will provide the framework for subsequent studies, which will be necessary to get a more complete sense of innovations in faith-based service provision and the changing nature of relationships between the government and faith-based organizations.

Indicator Construction

The analysis presented in this paper was based on a set of indicators that were constructed to permit the comparison of means between secular organizations and FBOs active in the low-income housing sector. Statistical comparison was based on an Independent Samples t-Test that compared the mean scores of two groups on a given variable. Statistical significance of observed difference was estimated based on the results of Levene's Test for Equality of Variances to assure the two groups had approximately equal variance on the dependent variable.

Structural characteristics, such as years in operation, total employees, total volunteers, and annual budget, were reported directly by responding organizations. Table A-4 reported this data. The activities of secular and faith-based housing nonprofits were determined based on organizational responses to whether they were involved in specific activities within four general areas: housing development, business development, social services, and “other.” Specific activities included the following:

• Housing development: rental housing, cooperative housing, home ownership, single room occupancy housing (SRO), elderly housing, housing for people with AIDS, housing for people with disabilities, single family housing, small residential, large residential, scattered site, land trust, Hope VI, lead paint abatement, and asbestos removal

128

• Business Development: commercial development, industrial development, incubator development, business operation, Brownfield redevelopment, small business loan program, small business technical assistance, micro enterprise development, and business district/main street revitalization

• Social Services: day care, summer camp, youth programs, drug abuse programs, teen/parent counseling, landlord/tenant mediation, crime prevention, cultural programs/centers, arts programs, voter registration, advocacy, home ownership counseling, leadership training, job counseling/job training

• Other services: mixed-use development, energy conservation, Community Reinvestment Act (CRA)

Counts of involvement were then made as follows, and are reported in Table A-5 and Figure A-1:

Housing production for secular and faith-based housing nonprofits was based on organizational responses to a set of questions about current and future projects with regard to the number of existing housing projects, number of units in each project, and total number of future (i.e., planned) units. The sum of these responses was reported in Table A-6. Organizations were also asked “what hurdles do you face as a nonprofit in the housing arena in building capacity and growing?” Open-ended responses were coded as follows: operating funds, administrative capacity, land acquisition, money, and “other.” An indicator was created for each coding category. When an organization cited a hurdle in a particular category, that indicator was given a value of 1. The mean of each indicator was interpreted as the proportion of organizations identifying a particular type of hurdle. These data are presented in Figure A-2 as impediments to growth and expanded organizational capacity.

All organizations were asked to list needed technical assistance in four areas: housing development, community and neighborhood planning, commercial development, and commercial strip revitalization. The open-ended data were coded as follows:

• Housing development: management, tax credits, real estate management, housing support, other

• Community and neighborhood planning: market studies, neighborhood planning, strategic planning, other

• Commercial development: financing, procedural-operational, staff development, other

• Commercial strip development: design assistance, staff development, business development, other

An indicator was created for each coding category. When an organization cited a technical assistance need in a particular category, that indicator was given a value of 1. The mean of each

129

indicator was interpreted as the proportion of organizations identifying a particular type of needed technical assistance. These data are reported in Figure A-3.

Organizations were asked to identify partners/collaborators for future housing development projects. Open-ended responses were coded into the following categories: banks, federal agencies, state agencies, county agencies, city agencies, individuals or private sector partners, and other partners. An indicator was created for each coding category. When an organization cited a type of partner for a particular category, that indicator was given a value of 1. The mean of each indicator was interpreted as the proportion of organizations identifying a particular type of partner.

Sources of financial problems were measured based on organizations’ responses to the survey question asking them to identify sources of financing that they were having difficulty accessing. Open-ended responses were coded into the following categories: foundations and corporations, federal agencies, state agencies, county agencies, and city agencies. An indicator was created for each coding category. When an organization cited a type of funding source, that indicator was given a value of 1. The mean of each indicator was interpreted as the proportion of organizations identifying a particular type of needed technical assistance. The data for partners and sources of financial problems are reported in Figure A-4. These indicators also form the basis of the tables presented in Appendix C.

APPENDIX C: SUPPLEMENTAL TABLES

Table C-1. Characteristics of Secular Organizations and Faith-Based Organizations

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 48 1986 12 1.7 Year Founded

Faith-based 27 1988 8 1.6 .45

Secular 48 21 12 1.7 Number of Years in Business Faith-based 27 19 8 1.6

.45

Secular 51 1 .000 .000 501(C)3 Status a

Faith-based 29 1 .18 .03 .33

(continued)

130

Table C-1. Characteristics of Secular Organizations and Faith-Based Organizations (continued)

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 42 $2,323,652 $3,434,243 $529,915 Annual Budget

Faith-based 26 $3,728,840 $13,217,195 $2,592,105 .51

Secular 48 1 .49 .07 CHDC Status a

Faith-based 29 1 .73 .14 .48

Secular 46 14.5 40 5.9 Total Employees

Faith-based 27 24.56 95 18.3 .53

Secular 42 8.76 21 3.2 Total Number of Full-time Employees Faith-based 24 4.08 5.4 1.1

.29

Secular 42 .12 .55 .085 Total Number of Volunteers Faith-based 24 .04 .2 .042

.51

Secular 42 4.4 15 2.4 Total Number of Part-time Employees a Faith-based 24 .92 1.1 .22

.15

Secular 37 4.9 10.9 1.8 Total Number of Employees with Benefits Faith-based 22 3.2 4.0 .8

.49

Secular 37 5.9 22 3.7 Total Number of Employees without Benefits Faith-based 21 .95 1 .24

.42

Secular 41 .0140 .07 .01 Ratio of Volunteers to Employed Workers Faith-based 23 .0017 .01 .001

.01***

Secular 41 2.2 1.4 .21 Total Number of Existing Housing Projects Faith-based 19 2.8 1.2 .28

.11

Secular 51 57 89 13 Total Number of Units in Existing Projects Faith-based 29 63 84 16

.77

Secular 51 34 62 9 Total Number of Units in Future Projects Faith-based 29 63 87 16

.09*

Secular 40 $9,456 $16,610 $2,626 Technical Assistance Budget a Faith-based 21 $24,810 $65,576 $14,310

.30

Secular 30 $20,095.00 $55,732.830 $10,175 Training Budget

Faith-based 20 $10,395.00 $26,525.390 $5931 .72

Secular 51 7.6078 3.68824 .51646 Total Number of Reported Activities Faith-based 29 6.2759 3.61442 .67118

.12

(continued)

131

Table C-1. Characteristics of Secular Organizations and Faith-Based Organizations (continued)

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 34 .51 .450 .077 Percentage of Legal Services Paid at Market Rate Faith-based 21 .25 .420 .092

.040**

Secular 31 .43 .426 .076 Percentage of Legal Services Paid at Reduced Rate Faith-based 20 .35 .414 .093

.535

Secular 29 .43 .405 .075 Percentage of Legal Services Provided Free Faith-based 25 .59 .435 .087

.162

Secular 40 $4,528.18 $6,707 $1,060.402 Annual Legal Budget a

Faith-based 22 $11,173.55 $22,241 $4,741.975 .19

*** Significant at the 0.01 level, ** Significant at the 0.05 level, * Significant at the 0.10 level. a t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance assumed.

Assumption used based on result of Levene's Test for Equality of Variances.

Table C-2. Housing Development Activities of Secular and Faith-Based Nonprofit Organizations

Organization Type N

Proportion Yes

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 51 .53 .504 .071 Rental Housing

Faith-based 29 .62 .494 .092 .435

Secular 51 .00 .000a .000 Cooperative Housing

Faith-based 29 .00 .000a .000 ___

Secular 51 .67 .476 .067 Homeownership

Faith-based 29 .66 .484 .090 .918

Secular 51 .06 .238 .033 SRO Housing

Faith-based 29 .07 .258 .048 .859

Secular 51 .24 .428 .060 Elderly Housing

Faith-based 29 .24 .435 .081 .952

Secular 51 .12 .325 .046 AIDS Housing

Faith-based 29 .14 .351 .065 .795

Secular 51 .24 .428 .060 Disability Housing b

Faith-based 29 .10 .310 .058 .117

(continued)

132

Table C-2. Housing Development Activities of Secular and Faith-Based Nonprofit Organizations (continued)

Organization Type N

Proportion Yes

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 51 .69 .469 .066 Single Family Housing

Faith-based 29 .62 .494 .092 .557

Secular 51 .29 .460 .064 Small Residential b

Faith-based 29 .17 .384 .071 .210

Secular 51 .16 .367 .051 Large Residential

Faith-based 29 .14 .351 .065 .822

Secular 51 .51 .505 .071 Scattered Site

Faith-based 29 .48 .509 .094 .819

Secular 51 .10 .300 .042 Land Trust

Faith-based 29 .07 .258 .048 .663

Secular 51 .02 .140 .020 Hope VI

Faith-based 29 .00 .000 .000 .454

Secular 51 .24 .428 .060 Lead Paint Abatement

Faith-based 29 .24 .435 .081 .952

Secular 51 .04 .196 .027 Asbestos Removal

Faith-based 29 .07 .258 .048 .563

Secular 51 .04 .196 .027 Condos to lease to own

Faith-based 29 .03 .186 .034 .094*

* Significant at the 0.10 level. a t cannot be computed because the standard deviations of both groups are 0. b Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

133

Table C-3. Proportion of Organizations Reporting Specific Social Service Provided, by Organization Type

Organization

Type N

Proportion Responding

Yes Std.

Deviation Std. Error

Mean t-Test

Significance

Secular 51 .16 .367 .051 Day Care

Faith-based 29 .17 .384 .858

Secular 51 .04 .196 .027 Summer Camp a

Faith-based 29 .10 .310 .058 .320

Secular 51 .20 .401 .056 Youth Programs

Faith-based 29 .28 .455 .084 .418

Secular 51 .04 .196 .027 Drug Abuse Counseling Faith-based 29 .07 .258 .048

.563

Secular 51 .02 .140 .020 Teen Parent Counselinga

Faith-based 29 .03 .186 .034 .709

Secular 51 .12 .325 .046 Landlord/Tennant Mediation Faith-based 29 .03 .186 .034

.211

Secular 51 .14 .348 .049 Crime Prevention Faith-based 29 .10 .310 .058

.665

Secular 51 .10 .300 .042 Cultural Programs/Centers Faith-based 29 .07 .258 .048

.663

Secular 51 .02 .140 .020 Arts Program Faith-based 29 .10 .310 .058

.101

Secular 51 .02 .140 .020 Voter Registration

Faith-based 29 .03 .186 .034 .687

a Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance assumed. Assumption used based on result of Levene's Test for Equality of Variances.

Table C-4. Proportion of Organizations Reporting Specific Business Development Services, by Organization Type

Organization

Type N

Proportion Responding

Yes Std.

Deviation Std. Error

Mean t-Test

Significance

Secular 51 .24 .428 .060 Commercial Development

Faith-based 29 .21 .412 .077 .773

Secular 51 .02 .140 .020 Industrial Development

Faith-based 29 .03 .186 .034 .687

(continued)

134

Table C-4. Proportion of Organizations Reporting Specific Business Development Services, by Organization Type (continued)

Organization

Type N

Proportion Responding

Yes Std.

Deviation Std. Error

Mean t-Test

Significance

Secular 51 .04 .196 .027 Incubator Development

Faith-based 29 .03 .186 .034 .916

Secular 51 .04 .196 .027 Operate a Businessa

Faith-based 29 .17 .384 .071 .090 *

Secular 51 .14 .348 .049 Brownfield Redevelopment

Faith-based 29 .17 .384 .071 .677

Secular 51 .08 .272 .038 Small Business Loan program

Faith-based 29 .03 .186 .034 .441

Secular 51 .12 .325 .046 Small Business Technical Assistance

Faith-based 29 .07 .258 .048 .492

Secular 51 .18 .385 .054 Micro Enterprise Developmenta

Faith-based 29 .03 .186 .034 .029**

Secular 51 .22 .415 .058 Business District/Main Street Development

Faith-based 29 .07 .258 .048 .055*

** Significant at the 0.05 level, * Significant at the 0.10 level a Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

Table C-5. Proportion of Organizations Reporting Other Services, by Organization Type

Organization Type N Proportion

Responding Yes Std.

Deviation Std. Error

Mean t-Test

Significancea

Secular 29 .14 .351 .065 Mixed Use

Faith-based 51 .18 .385 .054 .492

Secular 29 .24 .435 .081 Energy Conservation

Faith-based 51 .16 .367 .051 .822

Secular 29 .14 .351 .065 Community Reinvestment

Faith-based 51 .18 .385 .054 .658

a Equal variance assumed based on result of Levene's Test for Equality of Variances.

135

Table C-6. Impediments to Achieving Organization Missiona

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 51 .5686 .50020 .07004 Operating Funds

Faith-based 29 .6207 .49380 .09170 .654

Secular 51 .3333 .47610 .06667 Administrative Capacity Faith-based 29 .2759 .45486 .08447

.599

Secular 51 .2157 .41539 .05817 Land Acquisition

Faith-based 29 .2759 .45486 .08447 .549

Secular 51 .2549 .44014 .06163 Money

Faith-based 29 .3103 .47082 .08743 .599

Secular 51 .3137 .46862 .06562 Other

Faith-based 29 .3793 .49380 .09170 .557

a No statistically significant difference.

Table C-7. Training Needs in Housing

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Housing

Secular 51 .0980 .30033 .04205 Management

Faith-based 29 .0690 .25788 .04789 .663

Secular 51 .1373 .34754 .04867 Tax Credits a

Faith-based 29 .1379 .35093 .06517 .993

Secular 51 .0000 .00000 .00000 Real Estate Management

Faith-based 29 .0000 .00000 .00000

Cannot be computed

Secular 51 .0392 .19604 Housing Support a

Faith-based 29 .1379 .35093 .06517 .171

Secular 51 .2353 .42840 .05999 Other

Faith-based 29 .2414 .43549 .08087 .952

Community Planning

Secular 51 .0000 .00000 .00000 Management

Faith-based 29 .0000 .00000 .00000

Cannot be computed

Secular 51 .0980 .30033 .04205 Tax Credits

Faith-based 29 .0690 .25788 .04789 .663

(continued)

136

Table C-7. Training Needs in Housing (continued)

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Secular 51 .0392 .19604 .02745 Real Estate Management a

Faith-based 29 .0000 .00000 .00000 .159

Secular 51 .1176 .32540 .04556 Housing Support a

Faith-based 29 .2069 .41225 .07655 .321

Secular 51 .0000 .00000 .00000 Other

Faith-based 29 .0000 .00000 .00000

Cannot be computed

a Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance assumed. Assumption used based on result of Levene's Test for Equality of Variances.

Table C-8. Proportion of Organizations Identifying Needed Technical Assistance

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Housing

Secular 51 .1765 .38501 .05391 Management Total

Faith-based 29 .2414 .43549 .08087 .492

Secular 51 .1373 .34754 .04867 Tax Credits Total

Faith-based 29 .3103 .47082 .08743 .090*

Secular 51 .0000 .00000 .00000 Real Estate Management Total Faith-based 29 .0345 .18570 .03448

.326

Secular 51 .0784 .27152 .03802 Housing Support Total

Faith-based 29 .1034 .30993 .05755 .708

Secular 51 .3137 .54736 .07665 Other Total

Faith-based 29 .1724 .38443 .07139 .288

Community Planning

Secular 51 .0000 .00000 .00000 Market Studies a

Faith-based 29 .1034 .30993 .05755 .083*

Secular 51 .0980 .30033 .04205 Neighborhood Planning a

Faith-based 29 .2759 .45486 .08447 .066*

Secular 51 .0588 .23764 .03328 Strategic Planning

Faith-based 29 .0345 .18570 .03448 .613

Secular 51 .1176 .32540 .04556 Other

Faith-based 29 .0690 .25788 .04789 .464

(continued)

137

Table C-8. Proportion of Organizations Identifying Needed Technical Assistance (continued)

Organization Type N Mean

Std. Deviation

Std. Error Mean

t-Test Significance

Commercial Development

Secular 51 .0980 .30033 .04205 Financing a

Faith-based 29 .2069 .41225 .07655 .219

Secular 51 .0980 .30033 .04205 Procedural-Operational a

Faith-based 29 .2069 .41225 .07655 .219

Secular 51 .0980 .30033 .04205 Staff Development a

Faith-based 29 .2069 .41225 .07655 .219

Secular 51 .0980 .30033 .04205 Other a

Faith-based 29 .2069 .41225 .07655 .219

Commercial Strip Development

Secular 51 .0392 .19604 .02745 Design Assistance a

Faith-based 29 .0000 .00000 .00000 .159

Secular 51 .0196 .14003 .01961 Staff Development a

Faith-based 29 .0000 .00000 .00000 .454

Secular 51 .0392 .19604 .02745 Business Development a

Faith-based 29 .0000 .00000 .00000 .159

Secular 51 .0392 .19604 .02745 Other a

Faith-based 29 .0000 .00000 .00000 .159

* Significant at the 0.10 level a Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

Table C-9. Barriers to Receiving Needed Technical Assistance

Organization Type N Mean Std. Deviation

t-Test Significance

Secular 51 .1373 .34754 Funding a

Faith-based 29 .3103 .47082 .090*

Secular 51 .0588 .23764 Lack of Knowledge

Faith-based 29 .0690 .25788 .859

Secular 51 .0588 .23764 Lack of Time a

Faith-based 29 .1034 .30993 .505

(continued)

138

Table C-9. Barriers to Receiving Needed Technical Assistance (continued)

Organization Type N Mean Std. Deviation

t-Test Significance

Secular 51 .0784 .27152 Staff Capacity

Faith-based 29 .0000 .00000 .125

Secular 51 .1569 .36729 Other

Faith-based 29 .1724 .38443 .858

*Significant at the 0.10 level a Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

Table C-10. Perceived Barriers to Project Planning Process

Organization Type N Mean

Std. Deviation

T-Test Significance

Secular 51 .0588 .23764 Zoning a

Faith-based 29 .1379 .35093 .286

Secular 51 .1569 .36729 Lot size, availability, cost a

Faith-based 29 .1724 .38443 .860

Secular 51 .1569 .36729 Bureaucracy/Politics

Faith-based 29 .4138 .50123 .010*

Secular 51 .0784 .27152 Citizens/NIMBY a

Faith-based 29 .1034 .30993 .718

Secular 51 .0784 .27152 Ordinance problems a

Faith-based 29 .0000 .00000 .044**

Secular 51 .0588 .23764 Financial

Faith-based 29 .0000 .00000 .188

Secular 51 .0980 .30033 Other

Faith-based 29 .0690 .25788 .663

** Significant at the 0.05 level, * Significant at the 0.10 level a Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

139

Table C-11. Categories of Project Partners Identified, by Organization Type a

Organization Type N

Proportion of Organizations Identifying Partner Type

Std. Deviation

t-Test Significance

Secular 51 .04 .20 Banks b

Faith-based 29 .10 .31 .32

Secular 51 .10 .30 Federal Agencies

Faith-based 29 .07 .26 .66

Secular 51 .16 .37 State Agencies b

Faith-based 29 .07 .26 .22

Secular 51 .08 .27 County Agencies b

Faith-based 29 .00 .00 .04**

Secular 51 .16 .37 City Agencies

Faith-based 29 .10 .31 .51

Secular 51 .16 .37 Individuals or Private-Sector Partners Faith-based 29 .14 .35

.82

Secular 51 .08 .27 Other Partners b

Faith-based 29 .24 .44 .08*

** Significant at the 0.05 level, * Significant at the 0.10 level a Based on partners identified over a series of current and projected projects. Organizations are assigned a score of

1 for each partner type if they identify that partner type in any reported project. b Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

Table C-12. Sources of Financial Problems Identified, by Organization Type a

Organization Type N Mean

Std. Deviation

Std. Error Mean

T-Test Significance

Secular 51 .22 .42 .06 Foundations and Corporations Faith-based 29 .21 .41 .08

.928

Secular 51 .20 .40 .06 Federal Agencies

Faith-based 29 .21 .41 .08 .909

Secular 51 .18 .39 .05 State Agencies

Faith-based 29 .14 .35 .07 .658

Secular 51 .02 .14 .02 County Agencies b

Faith-based 29 .10 .31 .06 .177

Secular 51 .25 .44 .06 City Agencies b

Faith-based 29 .41 .50 .09 .161

a Organizations are assigned a score of 1 for each source of financial problem they identify. b Indicates t-test computed with group variance not assumed to be equal across groups. Otherwise equal variance

assumed. Assumption used based on result of Levene's Test for Equality of Variances.

140

REFERENCES

Babbie, E. (1990). Survey research methods. Belmont, CA: Wadsworth.

Berger, P. L., & Neuhaus, R. J. (1977). To empower people: The role of mediating structures in public policy. Washington, DC: American Enterprise Institute for Public Policy Research.

Birch, E. L., & Gardner, D. S. (1981). The seven percent solution: A review of philanthropic housing, 1870–1910. Journal of Urban History, 7, 403-438.

Bishop, C. M. (1991). Building on success: A report on state capacity-building programs targeted to nonprofit housing developers. Washington, DC: National Support Center for Low-Income Housing.

Bowen v. Kendrick 612 (487 U.S. 589, 612 1988).

Boyte, H. C., & Riessman, F. (1986). The new populism: The politics of empowerment. Philadelphia: Temple University Press.

Bratt, R. G., Keyes, L. C., Schwartz, A., & Vidal, A. C. (1994). Confronting the management challenge: affordable housing in the nonprofit sector. New York: Community Development Research Center.

Bratt, R. G., & Rohe, W. M. (2005). Challenges and dilemmas facing community development corporations in the United States. Community Development Journal, 42(1), 63–78.

Briggs, X. d. S. (2004). Faith and mortar: Religious organizations and affordable housing strategy in urban America. In R. V. Anglin (Ed.), Building the organizations that build communities (pp. 43-54). Washington, DC: U.S. Department of Housing and Urban Development.

Burt, M. R., Aron, L. Y., Douglas, T., Valente, J., Lee, E., & Iwen, B. (1999). Homelessness: Programs and the people they serve—summary report. Findings of the National Survey of Homeless Assistance Providers and Clients. Washington, DC: Urban Institute.

Community and Economic Development Association of Michigan (CEDAM). (2004). Standing strong: Rebuilding neighborhoods and revitalizing communities. 2004 CED accomplishments. Lansing, MI: Community and Economic Development Association of Michigan.

Chang, P., Williams, D., Griffith, E., & Young, J. (1994). Church-agency relationships in the black community. Nonprofit and Voluntary Sector Quarterly, 23, 107–118.

Chaves, M. (1999a). Congregation's social service activities. Policy brief. Washington DC: The Urban Institute.

Chaves, M. (1999b). Religious congregations and welfare reform: Who will take advantage of "Charitable Choice"? American Sociological Review, 64(6), 836–846.

141

Chaves, M. (2002). Religious organizations: Data resources and research opportunities. American Behavioral Scientist, 45(10), 1523–1549.

Chaves, M., & Tsitsos, W. (2001). Congregations and social services: What they do, how they do it, and with whom. Nonprofit and Voluntary Sector Quarterly, 30(4), 660–683.

Cisneros, H. G. (1996). Higher ground: Faith communities and community building. Washington DC: U.S. Department of Housing and Urban Development.

Clerkin, R. M., & Grønbjerg, K. A. (2007). The capacities and challenges of faith-based human service organizations. Public Administration Review, 67(1), 115–126.

Cnaan, R. A. (1997). Social and community involvement of religious congregations housed in historic religious properties: Findings from a six-city study. Final report to Partners for Sacred Places. Philadelphia: University of Pennsylvania.

Cohen, C. S., Mulroy, E., Tull, T., White, C., & Crowley, S. (2004). Housing plus services: Supporting vulnerable families in permanent housing. Child Welfare, 83(5), 509–528.

Crittenden, W. F. (2000). Spinning straw into gold: The tenuous strategy, funding, and financial performance linkage. Nonprofit and Voluntary Sector Quarterly, 29, 164–182.

de Toqueville, A. (1969). Democracy in America. New York: Harper and Row.

Devita, C., Platnick, T., & Twombly, E. (1999). Findings from the survey of community services of faith-based organizations in New Jersey. Washington, DC: Center for Nonprofits and Philanthropy, The Urban Institute.

DiMaggio, P. J., & Powell, W. W. (1991). The new institutionalism in organizational analysis. Chicago: University of Chicago Press.

Ebaugh, H. R., Chafetz, J. S., & Pipes, P. (2005a). Faith-based social service organizations and government funding: Data from a national survey. Social Science Quarterly, 86(2), 273–292.

Ebaugh, H. R., Chafetz, J. S., & Pipes, P. (2005b). Funding good works: Funding sources of faith-based social service coalitions. Nonprofit and Voluntary Sector Quarterly, 34(4), 448–472.

Ebaugh, H. R., Pipes, P., & Chafetz, J. S. (2006). Where’s the faith in faith-based organizations? Measures and correlates of religiosity in faith-based social service coalitions. Social Forces, 84(4), 2259–2272.

Ebaugh, H. R., Pipes, P., Chafetz, J. S., & Daniels, M. (2003). Where's the religion? Distinguishing faith-based from secular social service agencies. Journal for the Scientific Study of Religion, 42(3), 411–426.

142

Fannie Mae Foundation. (2001). Facts about the role of faith-based organizations in providing affordable housing and financial services in American communities. Washington, DC: Fannie Mae.

Garland, D. R. (1992). Church social work: Helping the whole person in the context of the church. Botsford, CT: North American Christians in Social Work.

Goetz, E. G. (1992). Local government support for nonprofit housing: A survey of U.S. cities. Urban Affairs Quarterly, 27(3), 420–435.

Gormley, W. T. (1991). Privatization and its alternatives. Madison, WI: University of Wisconsin Press.

Gronbjerg, K. A. (1993). Understanding nonprofit funding: Managing revenues in social services and community development organizations. San Francisco: Jossey-Bass.

Guo, C. (2007). When government becomes the principal philanthropist: The effects of public funding on patterns of nonprofit governance. Public Administration Review, 67(3), 458–473.

Hager, M. A., Wilson, S., Pollak, T. H., & Rooney, P. M. (2003). Response rates for mail surveys of nonprofit organizations: A review and empirical test. Nonprofit and Voluntary Sector Quarterly, 32(2), 252–267.

Hall, P. D. (1992). Inventing the nonprofit sector. Baltimore, MD: The Johns Hopkins University Press.

Hays, R. A. (2002). Habitat for Humanity: Building social capital through faith based service. Journal of Urban Affairs, 24, 247–269.

Heim, S. (1995). God’s long shot in the inner city: A vision of church-based economic development. Christian Century, July 5–12, 680–682.

Hodgkinson, V. A., & Lyman, R. W. (1989). The future of the nonprofit sector: Challenges, changes, and policy considerations. San Francisco: Jossey-Bass.

Hodgkinson, V. A., & Weitzman, M. S. (1993). From belief to commitment: The community service activities and finances of religious congregations in the United States. Washington, DC: Independent Sector.

Housing Plus Services Committee of the National Low Income Housing Coalition. (2002). Housing Plus Services Typology. Retrieved February 28, 2008, from http://www.nlihc.org/doc/typology.pdf

Housing Plus Services Committee of the National Low Income Housing Coalition. (2006). Housing Plus Services Principles for Program Design and Implementation. Retrieved February 28, 2008, from http://www.nlihc.org/doc/principles.pdf

143

Hula, R. C. (1988). Market-based public policy. London: Macmillian.

Hula, R. C., Jackson-Elmoore, C., & Reese, L. A. (2007). Mixing God’s work and the public business: A framework for the analysis of faith-based service delivery. Review of Policy Research, 24(1), 67–89.

Jackson, C. Y. (1994). Production and financing choices for municipal services: Contracting and franchising. Los Angeles: University of Southern California.

Keating, D. W., Rasey, K. P., & Krumholz, N. (1990). Community development corporations in the United States: Their role in housing and urban redevelopment. In W. Vliet & J. Weesep (Eds.), Government and housing: Developments in seven countries (pp. 206–218). Newbury Park, CA: Sage.

Keyes, L. C. (1971). The role of nonprofit sponsors in the production of housing. Papers submitted to U.S. House Committee on Banking and Currency; Subcommittee on Housing Panels on Housing Production, Housing Demand, and Developing a Suitable Living Environment, 92nd Cong., 1st Sess., Pt. 1, 159-181.

La Barbera, P. (1992). Enterprise in religious-based organizations. Nonprofit and Voluntary Sector Quarterly, 21, 51–67.

Lincoln, C., & Mamiya, L. (1990). The black church in the African American experience. Durham, NC: Duke University Press.

Mayer, N. S. (1984). Neighborhood organizations and community development: Making revitalization work. Washington, DC: Urban Institute Press.

Mitchell, R. E., Florin, P., & Stevenson, J. F. (2002). Supporting community-based prevention and health promotion initiatives: Developing effective technical assistance systems. Health Education & Behavior, 29(5), 620–639.

Monsma, S. V. (1996). When sacred and secular mix: Religious nonprofit organizations and public money. Lanham, MD.: Rowman & Littlefield.

Monsma, S. V. (2002). Faith-based organization as partners with government in providing welfare to work services. Paper presented at the American Political Science Association Annual Conference, Boston.

National Alliance of Community Economic Development Associations. (2007). Community development fact sheet. Arlngton, VA: Author.

National Commission on Civic Renewal. (1998). A nation of spectators: How civic disengagement weakens America and what we can do about it. College Park, MD: The National Commission on Civil Renewal.

National Congress for Community Economic Development. (1999). Coming of age: Trends and achievements of community-based development organizations. Washington, DC: Author

144

National Congress for Community Economic Development. (2005). Reaching new heights. Trends and achievements of community-based development organizations: 5th National Community Development Census. Washington, DC: Author

National Performance Review. (1993). Creating a government that works better and costs less: Report of the National Performance Review. The Gore Report on Reinventing Government. Washington DC: Times Books.

Netting, F. E. (1984). Church-related agencies and social welfare. Social Service Review, September, 404–420.

Printz Platnick, T. J. (1998). Faith-based service providers in the nation's capitol: Can they do more? Public policy brief, charting civil society, no. 2, April. Washington, DC: The Urban Institute.

Reese, L., & Shields, G. (1999). Economic development activities of urban religious institutions. International Journal of Economic Development, 2, 165–199.

Reese, L. A., & Shields, G. (2000). Faith-based economic development. Policy Studies Review, 17, 84–103.

Roozen, D., McKinney, W., & Carroll, J. (1984). Varieties of religious presence: Mission in public life. New York: Pilgrim Press.

Rubin, H. J. (1993). Understanding the ethos of community-based development: Ethnographic description for public administrators. Public Administration Review, 53(5), 428–437.

Salamon, L. M. (1987a). Of market failure, government failure, and third-party government: Toward a theory of government-nonprofit relations in the modern welfare state. Journal of Voluntary Action Research, 16(1–2), 29–49.

Salamon, L. M. (1987b). Partners in public service: The scope and theory of government-nonprofit relations. In W. W. Powell (Ed.), The nonprofit sector: A research handbook (pp. 99-117). New Haven, CT: Yale University Press.

Salamon, L. M. (1998). America's nonprofit sector: A primer (2nd ed.). New York: The Foundation Center.

Savas, E. S. (2000). Privatization and public-private partnerships. New York: Chatham House.

Smith, S. R., & Lipsky, M. (1993). Nonprofits for hire: The welfare state in the age of contracting. Cambridge, MA: Harvard University Press.

Smith, S. R., & Sosin, M. R. (2001). The varieties of faith-related agencies. Public Administration Review, 61(6), 651–670.

Twombly, E. (2002). Religious versus secular human service organizations: Implications for public policy. Social Science Quarterly, 83(4), 947–961.

145

VanTil, J. (1988). Mapping the third sector: Voluntarism in a changing social economy. New York: Foundation Center.

Vidal, A. C. (1992). Rebuilding communities: A national study of urban community development corporations. New York: New School for Social Research, Community Development Research Center.

Vidal, A. C. (1995). Reintegrating disadvantaged communities into the fabric of urban life: The role of community development. Housing Policy Debate, 6, 169–230.

Walker, C. (1993). Nonprofit housing development: Status, trends, prospects. Housing Policy Debate, 4(3), 339–414.

Wineburg, R. J. (1992). Local human services provision by religious congregations: A community analysis. Nonprofit and Voluntary Sector Quarterly, 23, 107–118.

Wineburg, R. J. (1993). Social policy, community service development, and religious organizations. Nonprofit Management and Leadership, 3(3), 283–297.

Wineburg, R. J. (1996). An investigation of religious support of public and private agencies in one community in an era of retrenchment. Journal of Community Practice, 3(2), 35–56.

146

OPENING THE BLACK BOX OF FAITH-BASED SERVICES: MEASURING PROGRAM COMPONENTS AND TREATMENT DOSE

Fred J. De Jong Claudia Horn

What is the significance of faith in substance abuse recovery programs? How is faith used and in what ways does it help a person recover from addiction? Is faith in God, relationships with others, or both the way to a clean, sober, and more responsible lifestyle? These are some of the questions that inspired 17 Gospel Rescue Missions (GRMs)—faith-based providers of substance abuse treatment and recovery support programs that rarely receive state or federal funding—from around the country to participate in a pilot study to help translate faith into measurable research indicators. This research involved working with the leadership and staff of these GRMs to engage in a process to help define and measure the faith-motivated help provided to homeless and drug-dependent persons. The aim of this research was to document the nature of the GRMs’ work and to lay the foundation for testing program effectiveness.

The role that faith plays in faith-based organizational settings is not well understood (Johnson, 2002). Thus, there is a need to better understand how faith is defined at the organizational level in terms of services and interactions with clients and at the client level in regard to how individual faith is related to behavioral outcomes (Jeavons, 1994). This paper reports on a unique pilot study that examined the faith-based service provision of the GRMs.

To develop measures of the components of faith that permeate the GRMs at the organizational and service levels, work groups were convened composed of various staff members and interviews conducted with key stakeholders about the organizational mission and services. The information derived from these activities resulted in a set of initial measures of organizational faith that were piloted through a Web-based survey with the 17 GRMs. This paper reports on the results of the pilot survey measuring organizational faith, client faith, and the interactions between providers and clients. Follow-on research will focus on examining the effects of the program faith components on client outcomes, such as substance abuse and recovery.

A MIXED PUBLIC-PRIVATE SYSTEM OF CARE

When examining the role of faith in social service programs, it is helpful to review the history of and role that faith has played in the development of social service provision. The founders of our nation discussed the importance of “public religion,” as Ben Franklin called it, to promote moral

147

consensus and improve public life. “The great news about America,” according to Jon Meacham (2006) “is that religion shapes the life of the nation without strangling it.… The balance between the promise of the Declaration of Independence, with its evocation of divine origins and destiny, and the practicalities of the Constitution, with its checks on extremism, remains perhaps the most brilliant American success” (p.5).

Given this context, it is not surprising that religious institutions have long played an important role in social service delivery in the United States. For example, the colonists responded to the needs in their communities in the ways they had been taught and had observed in their homelands, where churches and synagogues took the lead in providing communal support for the poor ever since the breakup of Europe’s feudal estates (Specht & Courtney, 1994). Settlement houses, the message of the social gospel, and a multitude of charity organizations—from the international reach of the Salvation Army to the tiny one-person ministry that serves a neighborhood’s troubled youths—represent the diverse influences that shape the present-day response to those in need (Ellor, Netting, & Thibault, 1999).

The historical result of these efforts is a mixed public-private human services delivery system that depends largely on federal resources, is often administered by states, and sustains many thousands of local voluntary organizations with programs to serve populations in need (Karger & Stoesz, 2005). Most localities offer publicly delivered services, such as child protective services, and contract with private, community-based providers, many of which are faith-based (Ambrosino, Heffernan, Shuttleworth, & Ambrosino, 2001). Local entities providing publicly funded services often represent a variety of religious motivations or may have no religious orientation at all, such as advocacy organizations for the disabled, local urban leagues, associations for the blind, labor unions, and fraternal organizations (Karger & Stoesz, 2005).

A Changing Service Delivery Environment

Recently, there are new opportunities for faith-based organizations, particularly small organizations and congregations, to compete with secular organizations for federal grants and support, while maintaining their religious identity (Cnaan & Boddie, 2002). The Charitable Choice provisions of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 and the current Centers for Faith-Based and Community Initiatives in 10 major federal agencies encourage faith-based and community-based initiatives. The Bush administration has recognized the trust much of the public places in the role of faith and faith-related organizations to facilitate individual change in program participants (Boddie and Cnaan, 2006; Smith & Sosin, 2001). Moreover, many people believe that persons with religious commitment “go above and

148

beyond the call of duty with clients or act in ways that inspire an unusual degree of trust among program beneficiaries” (Dilulio, 2002, p.56).

BUILDING AN EMPIRICAL BASE FOR FAITH-BASED SERVICE DELIVERY

Although public policy and public discourse may support faith-based and community-based initiatives, there is little empirical evidence to support the relative effectiveness of faith-based versus secular service delivery. Johnson (2002) noted that although much has been written about faith-based interventions, very few sources specify the exact elements of the intervention. If faith-related services are to compete for mainstream resources, decision-makers will require evidence of effectiveness. The standards for testing effectiveness are well-known, but they are rarely applied to the majority of human service interventions (Johnson, 2002). Thus, more rigorous evaluation studies are essential to gain transparency in the “appropriate management and design of social service in specific settings for particular groups of clients” (von Furstenberg, 2006, p.58).

The Current Literature

The literature presents an emerging body of empirical findings related to faith-based social service delivery, welfare-to-work programs, youth and parent education models, transitional housing, and substance abuse treatment. However, faith-based service evaluation is in its infancy (Boddie & Cnaan, 2006). Overall, the extant research has found that faith-based services are neither inferior nor superior to services provided by secular organizations.

Boddie and Cnaan (2006) recommended key “next steps” for developing the knowledge base on faith-based services, such as applying social science methods to help open the “black box” of faith-based service delivery to increase understanding of the services that clients actually receive and how these services make a difference in their lives. Specifically, Boddie and Cnaan suggest research to carefully measure the extent of faith-based program elements in a particular service and to track participant exposure to faith components. Outcome measures should be specific enough to link the faith-based aspects of a service to client change over time. Grettenberger, Bartkowski, and Smith (2006) further recommended that future research use secular comparison groups to help determine the differential effect of faith-based services, and warned against selection biases that may funnel only some types of clients, such as the most religious or the most motivated, to faith-based services. An additional key research concern is sample size (Boddie & Cnaan, 2006; Grettenberger et al., 2006). Thus, future research should recruit larger

149

and more representative samples of clients and take steps to limit the high attrition or loss of study participants during the course of a study.

CONCEPTUAL CLARITY: MEASURING PROGRAMMATIC FAITH

Measurements of faith can benefit from a history of testing as well as several well-established scales, such as the Faith Maturity Scale (Benson, Donahue, & Erickson, 1993). Research is beginning to test ways of measuring faith and its role in individual health outcomes. Koening, McCullough, and Larson (2007) reported that the connection between faith and various health and mental health variables has been widely studied. Fowler’s (1981) stages of faith development provide a helpful way to measure the function of faith through the life span. On a clinical level, social work practitioners use spiritual assessment as a way to evaluate their clients (Hodge, 2001).

However, faith in the context of program service delivery requires new ways to measure its effect on social service delivery models. Positive outcomes cannot be replicated if a study does not identify the type of faith components delivered, how much of this component was received (i.e., the dose), and the mode of delivery (e.g., group treatment, counseling, nonstaff lay mentoring). As in other human service interventions, a goal for faith-based program evaluation is to provide clear and systematic measurement of both the helping process and expected individual outcomes. To accomplish this, it is necessary to understand what faith means and how it is applied in social service programs.

A program’s faith elements and the manner in which they are communicated are referred to as “programmatic faith.” DiIulio (2002) views programmatic faith as the expression of organizational religiosity through a variety of program components. Programmatic faith can be a measurable component of a service, with its own measures of activities and related outcomes. Individual change through a faith-based service program could be seen as the product or result of a program’s “value-expressive activity” on behalf of its organization’s mission (Jeavons, 1994).

Defining Faith

Researchers have just begun to develop specific measures of programmatic faith; however, it is clear from the literature that there are no “one size fits all” measures of faith. The definition of faith may vary depending on the program and organizational context. One approach developed by Unruh and Sider (2005) delineates nine elements of faith-based programs with which to measure programmatic faith: religious references in program self-descriptions, objects with

150

religious associations in the program environment, invitations to a religious service or activity, use of prayer in service provision, use of sacred text in service provision, worship, sharing of personal testimonies, religious teachings or discussions, and invitations to a personal commitment to faith or spiritual renewal. Although Unruh and Sider’s conception of programmatic faith more closely inspects faith-related program elements, it mixes program descriptions, service environment measures, curricular content, modes of worship, and personal activities particular to only some faith traditions.

Hugen, De Jong, and Venema (2005) added clarity to Unruh and Sider’s approach by developing a distinct set of countable measures of organized, faith-related program services. They developed three distinct measures of programmatic faith: the centrality of spirituality within a program, the manner in which a program communicates its faith content to those it serves, and the type and frequency of faith-related program elements matched to a particular faith tradition. These three elements guide the development of measures capturing the faith framework of organizations. The next step in the measurement process is to make these measures specific to the organization’s service components and religious elements.

IDENTIFYING A PROGRAM’S FAITH FRAMEWORK

Selecting ways to measure a program’s faith-related components requires an intimate knowledge of the program’s faith framework. Valid measures will reflect the meaning and intended effect of each of the program’s components, which originate in the organization’s system of beliefs and resulting program choices.

In human services organizations in particular, organizations use implicit moral theory about human needs to select the appropriate problems their organization seeks to address. Hazenfeld (1983) noted that an organization’s set of beliefs present “desired moral values to be upheld and achieved when responding to these needs and problems, and the acceptable modes of response…. Patterns of service delivery will conform to these ideologies” (p. 100).

Consequently, one could expect organizational belief systems to be especially strong in faith-saturated service environments, such as the GRMs, where the organization’s faith content is believed to be relevant to the whole range of organizational characteristics. The GRMs’ beliefs will shape their mission, goals, structure, decision-making process, staff selection criteria, “change technology” or choice of intervention for participant change, manner of service delivery, and desired client outcomes (Jeavons, 1994). These authors suggest that the sum of these beliefs, value choices, and faith-shaped preferences constitute a faith framework for a program of

151

service. The impact of different religious traditions on such faith frameworks in organizations remains an undeveloped aspect of the literature, as one can expect that Buddhist, Muslim, or Jewish conceptions, for example, will differ dramatically in aspects of their faith-shaped program components, perceived essential services, the value placed on individual spiritual commitment and beliefs, and conceptualizations of how individuals change.

TESTING PROGRAMMATIC FAITH IN SUBSTANCE ABUSE TREATMENT

Substance abuse treatment provision has several qualities that make it a good modality of service in which to test faith-based services. First, substance abuse treatment programs are common in nearly every urban area in the United States, which would facilitate research access to large and varied samples in all regions. Second, most substance abuse treatment providers target similar high-risk, low-income populations as public treatment programs, making the recruitment of conventional treatment groups easier and offering broader applicability to findings. Third, these programs often entail a strong spiritual element, offering in most samples the complete range of faith-related measures and useful comparisons of extreme scores between faith- and nonfaith-related tests. Overall, rigorous research of faith-based substance abuse rehabilitation efforts affords the potential to “provide appropriate information to further practice and policy development... [because] successful recovery, one that lasts for a significant period of time, includes some form of personal transformation and change in lifestyle” (Zannis & Cnaan, 2006, p. 90).

The development of defensible measures of faith-based program components and outcomes would lay the foundation for comparative effectiveness studies with secular substance abuse treatment providers. The results could guide public policy by identifying which elements of faith-based treatment programs show the greatest effect for positive outcomes and which treatment populations show the greatest benefit.

The literature documents the widespread use of spiritual and religious support groups as a means for treating addictive diseases (Chen, 2006; Ellis & Schoenfeld, 1990; Koenig, 1998; Warfield & Goldstein, 1996). Since the founding of Alcoholics Anonymous (AA) in 1935, over a million members have sought recovery from substance abuse through its emphasis on spirituality, peer support, and accountability (Pardini Plante, Sherman, & Stump, 2000). In addition, “a vast majority of [substance abuse] treatment programs are based on the spiritual program of AA” (Pardini et al., 2000, p. 348).

152

Although success rates are low in substance abuse treatment in general, the chances of successful recovery may be enhanced by conscious and deliberate use of programmatic faith components (Miller, 1997). Some individuals may find sufficient strength in religion and spirituality to abstain or to reduce their substance use and improve their quality of life (Gorsuch, 1995). Other research sees positive effects for substance abuse treatment participants if they focus on developing a spiritual understanding as well as participating in a religion (Koenig et al., 2001). Although many authors theorize about how spirituality could be effective in substance abuse treatment, little empirical research analyzes how these spiritually driven programs assist individual transformation or how to measure programmatic impact on individual outcomes (Pardini et al., 2000).

To study the effect of programmatic faith in substance abuse treatment programs, the GRMs were selected because they make up an affiliation of organizations whose faith framework for service is well developed and permeates nearly all aspects of GRM substance abuse recovery programs. A review of GRM mission statements, goal areas and choice of service objectives, decision-making processes, staff selection, the type and manner of mission services (their change technology), and the range of desired participant outcomes demonstrates the strong influence of organizational beliefs and value preferences.

METHODS

Sample Description

Across the United States, 20 substance abuse treatment programs within 17 GRMs participated in the National Recovery Initiative (NRI) Pilot Study (see Appendix A). NRI included participating GRMs in 12 states—Northeast (ME, NY, PA), South (GA, NC, SC), Southwest (AZ, CA), Northwest (ID), and Midwest (IN, OH, TN)—and the District of Columbia.

GRM Background

Although “the primary purpose of GRMs is to save lost souls” (Hertel, 1999), the GRMs are major providers of faith-based substance abuse treatment and recovery support programs. Refusing state or federal funding for most of their programmatic work, the GRMs’ approach to solving substance abuse problems hinges on the application of strong Christian values placed on personal, faith-driven transformation (i.e., being “born-again”). Their faith framework drives a commitment to improving the lives of the urban poor, the homeless, and the chemically addicted through providing food, shelter, clothing, safety, health care, and family aid. The Association for Gospel Rescue Missions (AGRM) now represents over 300 missions around the world. Almost

153

100 years since its inception, the AGRM, of which the NRI pilot GRMs are members, is one of the 10 largest nonprofit human service systems in the country.

Data Collection Experience

AGRM members often have little experience in data collection. For example, many AGRM members use voluntary self-report surveys and questionnaires, as opposed to regularly scheduled service and results-based reports. Further, they often rely on data collected by other professional vendors and governmental programs. For instance, a study called the “Snapshot Survey of the Homeless” is compiled annually, wherein current data are compared with data collected in the previous year. Data are collected on about 20,000 participants representing 150 missions. These data share some similarities with the pilot evaluation in that they represent data collection in an environment usually not experienced in data collection for tracking service delivery and client outcomes.

Faith Framework

The GRM faith framework—the faith principles and philosophy that guide the GRMs’ work and service delivery choices—could be characterized as a fundamentalist, Protestant Christian framework. Similar to other faith-based organizations, the GRMs view substance abuse as a “life-controlling” circumstance that requires a holistic recovery approach (Ebaugh, Chafetz, & Pipes, 2006; Neff & McMaster, 2005; Sider & Unruh, 2004). Faith-based programs, such as the GRMs, seek to change the views and beliefs of participants in order to form “religious coping strategies” (Neff & MacMaster, 2005).

Developing Measures of Programmatic Faith in the GRM Context

The NRI procured financial support from two foundations to develop a pilot study for measurement development for substance abuse treatment programs in a GRM setting. During two week-long meetings, GRM leadership and management staff convened four work groups to articulate their faith framework into valid measures. The outcome of the first work group was a services received checklist for use by GRM participants in a substance abuse recovery program and as a GRM service taxonomy, including definitions and units of measurement. These were developed for two areas: measures of program service activities that are not heavily faith-based, such as substance abuse education classes, recreation, and skill development; and program service activities and participant outcomes that are heavily faith-based, such as one-on-one counseling, group therapy, spiritual instruction, and communal worship. The Service Activity Checklist form is completed daily by program participants to measure the program dose of faith-based and secular services.

154

The second GRM work group chose measures for non-faith-related outcomes of substance abuse residential treatment, indicators of recovery, and the data collection method. The third work group developed measures of participation in mission-sanctioned, therapeutic relationships with staff and mentors while in a recovery program.

The fourth GRM work group identified a set of faith indicators with multiple applications. In contrast to an organizational faith framework, which focuses on faith-driven choices around organizational variables, these measures of faith targeted the faith status of program participants and were intended to operationalize a personal faith framework as viewed by the GRMs. The indicators include self-identification by “being born-again” and several measures of faith status (faith maturity, spiritual growth measures, faith practices, and personal qualities affected by faith). These measures are an important part of the GRM pretest at intake and posttest at discharge, constituting an important personal faith change outcome for the GRM recovery programs.

By the conclusion of this work, four sets of reliable measures for a GRM faith framework had been generated. An extensive substance abuse residential treatment literature review contributed standardized forms of many of the non-faith-based service activities and treatment outcomes.

Draft hard-copy data collection instruments were then developed and circulated for comment, first to the NRI Steering Committee and then to all NRI member missions. After revisions, the draft hard-copy forms were reviewed by expert key informants drawn from GRM staff. On-site pretests at three GRMs (in Indiana, Ohio, and Tennessee) led to further refinements in the measurements and data collection protocol.

The final hard-copy data collection instruments were translated to Web-page presentation for centralized data collection and database management based on data input from Internet-connected work stations at each mission. Training on data input and Web page utilization was conducted via conference call, with extensive phone-based technical assistance throughout the implementation phase. The pilot survey covered the period from September 2007 to February 2008. When administered one-to-one by staff, the interview requires approximately 30 to 40 minutes.

NRI Pilot Study Measures

The complete set of NRI Pilot measures includes intake variables, such as client demographic information, baseline substance use and recovery indicators, a service activities checklist, measures of relationships between clients and staff, and faith status indicators. The service

155

activities checklist (see Appendix B) creates a cumulative record of service type and units of service received measured in hours. This measure is considered the program dose of services. The checklist is updated daily by clients.

A second set of indicators (see Appendix D), measures clients’ individual faith. These include four scales measuring (1) faith maturity, (2) spiritual growth, (3) faith-related activities, and (4) personal qualities. Another important faith measure includes one question about whether clients are “born-again.”

Details of the last set of measures regarding the relationships between clients and staff and clients’ ratings of their individual skills and self-esteem are also provided in Appendix D. The two measures of relationships include (1) the Working Alliance Inventory (WAI) (Horvath & Greenberg, 1989) and (2) Trust Staff Scale developed and tested with this project’s data. The two measures of an individual’s psychological adjustment include (1) Lovejoy’s Life Skills Scale (1995), and (2) Rosenberg’s Self-Esteem Scale (1965). The individual faith, relationships, and client’s psychological adjustment scales are administered at intake and then every 30 days after intake for a maximum of four repeated measures of each scale.

FINDINGS

Intake data were collected from 597 participants between the ages of 18 and 65, with a mean age of 40. About half of the participants provided demographic information at intake. Among those, the largest racial groups are White (53%) and African-American (42%). The GRM population is largely male (90%) and single (51%). About one in three has less than a high-school education (36%), and another one in three concluded their education with high-school graduation (37%). At intake, only one in four was employed full time. The participants’ principal drugs of choice are alcohol (36%), cocaine (27%), and marijuana (16%).

Type and Amount of Faith-Based Services Received

A Services Received Checklist serves as a measure of each participant’s experience with the GRMs’ faith-based program components. This is a daily record of services received, as recorded by participants. Service utilization data were submitted by only a portion of the GRMs participating in the NRI Pilot, which explains the variable sample size per measure. Another complicating factor is that program participants’ service participation data were counted during the study period, although the clients had different enrollment dates throughout the study. Therefore, exposure to program services varies by client.

156

Despite these limitations, there was notable variation in participant receipt of the faith-based program components and the program dose (see Table C-1 in Appendix C). Mission-related supervised work is by far the most common faith-based program component received, followed by chapel/worship and then resident-initiated spiritual activity, although there is large variation in the number of service units reported per participant for the same time period. Clients had low participation in individualized services, such as counseling or pastoral counseling. Figure C-1 (see Appendix C) shows that the majority of clients participated in five or more of the nine services that comprise the faith-related program elements. Only 12% of clients received one to four components, whereas 35% received five to six services, 29% received seven services, and 24% received eight to nine service components. Figure C-2 shows wider variation in the amount of faith-based services received in hours. Because it is unclear if the variation in program dose is due to participants’ enrollment date, this will be an important control variable in future analysis.

Individual Faith Indicators

Examining whether individual faith and spiritual growth indicators increased throughout the study are important outcome measures for the GRM participants. One of the goals of the GRMs is to contribute to the spiritual growth of the clients they serve. Five measures of faith status track the individual’s process of change from intake to discharge or follow-up: the Faith Maturity Scale, “born- again” self-report, a spiritual growth scale score, the faith activities scale, and the faith-related personal qualities scale.

Another faith measure marks spiritual growth by agreement with seven principal GRM beliefs. Table C-2 (see Appendix C) reports the mean responses to each of the seven questions included on the scale for approximately 350 clients. Clients were asked these questions at intake and every month during the pilot period. Because the responses included in this table represent clients’ last response to the scale, these results are not considered as pretest or posttest measures of individualized faith. Data for this measure indicate that clients have strong agreement in belief and intention. While it is evident from Table C-2 that clients participating in the pilot study experience, on average, strong spiritual growth, they rate their ability to apply the biblical principles in their lives much lower.

Another faith status measure uses self-report to ask, “Do you have a personal relationship with Christ (e.g., are you born-again)?” In results not shown in the tables, nearly half (48%) of participants responded yes (n=285). Because almost half of participants considered themselves born-again at intake, this finding may indicate a significant selection or social desirability bias in

157

which half of the clientele coming into the GRMs already identify with the mission faith framework

Measures of faith-related activities presented in Table C-3 (see Appendix C) vary significantly. While solitary prayer is the most frequently practiced activity, the range of responses varies. Attending worship services and personal devotions are the next most frequently cited activities. The GRM clients found it difficult to demonstrate biblical stewardship of money/income. This is not surprising given the low incomes of the clients served by GRMs.

The 15-item short version of the Faith Maturity Scale (Benson et al., 1993) provides a standardized measure of faith status at intake as well as change over time. The scale conceptualizes faith as a combination of vertical (the relationship between self and transcendent reality) and horizontal (the relationship of faith to serving humanity, mercy, and justice) dimensions of faith. In the GRM faith framework, consistent demonstration of certain attitudes and behaviors shows the opposite of characteristic addictive attitudes and behavior, indicating the progress of an internal change process. The understanding of spirituality in 12-step programs shows a similar conception; that is, spiritual change equals being honest, open-minded, and willing to change. Table C-4 (see Appendix C) indicates that, overall, the clients served by the GRMs have high ratings of faith maturity; however, there is some variation, with higher scores in helpfulness, gratitude and respect for others, and lower scores for the use of offensive language.

Relationship Building as a Faith-Based Services Mediating Variable

GRM residential substance abuse programs provide longer-term treatment than in most publicly supported settings. Among the GRM substance abuse treatment programs participating in the NRI Pilot, between 6 and 13 months is devoted to treatment, whereas many publicly supported residential substance abuse treatment programs provide 28 days or fewer of treatment (Neff, Shorkey, & Windsor, 2006). The GRM staff report frequent, therapeutically beneficial relationships developing between staff and participants or between program-sanctioned mentors and participants.

Several relationship measures track participant exposure to relationship components of GRM substance abuse treatment programs. Four composite measures gauge the strength of variables that impact relationships, three of which are commonly used, published scales: the quality and strength of a participant’s relationship with the staff person most influential in their recovery, as measured by the Working Alliance Inventory (Horvath & Greenberg, 1989); the self-rating Lovejoy Scale of Relationship Skills (Lovejoy, 1995); and Rosenberg’s Self-esteem Scale

158

(Rosenberg, 1965). Notably, the GRM clients rate their life skills and self-esteem as quite low on average. The fourth measure is the Trusting Staff Scale (see Appendix D), which measures the degree of trust placed in treatment staff. Analysis repeatedly shows statistically significant correlations between relationships while in treatment and the several measures of faith status. Table C-5 (see Appendix C) indicates that clients developed high levels of trust with GRM staff. Examining the correlations between worker trust and indicators of individual faith reveals that clients’ faith is positively associated with the development of relationships with the staff.

DISCUSSION

The NRI Pilot Study lays a preliminary foundation for more rigorous evaluation research by highlighting the need for greater conceptual clarity, specificity, and operationalization of key measures of programmatic faith. The answer to how faith program components actually contribute to individual outcomes begins to be found in greater specification of the faith-conditioned service or treatment. Inside the black box of faith-based service delivery, not all faith-based services are the same. In fact, faith-based service interventions and treatments vary considerably depending on the choices inherent in their faith frameworks as well as the specific treatment modality used by a program.

Moving forward, greater attention should be focused on understanding how the ideologies, beliefs, and inherent value preferences of a faith framework, both as applied to the client as well as the organization, can influence the kinds of services offered, the target population served, service delivery choices, and the types of participant outcomes valued by the faith-based providers. Organizational theory literature may help explain more specifically how a GRM faith framework permeates its service setting and the participant treatment experience. A systematic analysis of faith framework dimensions and their measures could potentially buttress the validity of faith-based service measures. Alternate faith frameworks, particularly non-Christian and those of the major world religions, deserve equivalent explication.

Effectiveness studies will depend on very careful examination of the aspects of a faith framework influencing a program’s service-related preferences. In fact, the kind of care taken in cross-cultural research, with its repeated checks on validity and partnering with indigenous experts, should be applied as well to the faith frameworks that shape faith-based service delivery.

The great variation in participant exposure to services in a faith-based environment should not be surprising. Similarly, variation in participant experience could be found in many non-faith-based, nonprofit service organizations. The range of service types received and the extent of their

159

receipt constitute the participant’s dose of faith-based treatment. How much intervention, for how long, delivered in what manner, and to whom are legitimate questions asked in human services effectiveness research. Faith-based service provision requires the same level of scrutiny.

The NRI Pilot Study puts flesh on the bones of programmatic faith and depicts a way to operationalize the faith content present in program components. While the precise substantive content of faith-based program components may vary by faith framework, programmatic faith is a useful concept that can connect the sphere of belief and values with the realm of implementation. It is here where the provider’s beliefs, values, and preferences translate into program components intended to shape an individual’s change process. Jeavons (1998) was correct when he conceived of faith-based service as the “value expression” of organizational mission.

The description of the GRM sample had an unexpected quality: about half of the participants— often residents of the most beleaguered parts of central cities and beset with addiction—share the mission faith framework. If the mission of the GRMs is to reach “the lost,” then about half of those they serve are not part of this target population. The GRM goal of individual faith transformation appears to have already been achieved prior to intake for this proportion of participants. Serious effort should be directed at measuring the degree of social desirability response in participant self-report; that is, respondents may provide information they think their interviewer wants to hear. However, self-selection into treatment settings is a common phenomenon. Service participants, when given the choice, will gravitate to the service setting that appeals to their backgrounds, or perhaps represents a familiar set of beliefs or shared values. The threat to an effectiveness study arises when such consonance of faith and value disproportionately sorts into faith-based programs those participants better equipped with coping skills or those with a higher level of motivation or greater access to social support and other recovery resources. Randomized assignment to treatment and comparison groups would help to mitigate such bias, and quasi-experimental matched comparison group designs can help fill the gap when randomization cannot be applied.

Programmatic faith may not drive all of the participant change process. The consistently significant and positive correlations between relationship aspects of treatment and the several measures of participant faith status suggest an interaction of relationship factors and faith status indicators. Longitudinal testing could remove the ambiguity of causal direction between relationship assets and faith status.

160

The greater duration of GRM substance abuse residential treatment programs provides the opportunity for therapeutic and mentoring relationships to develop and perhaps to mediate the participant service experience. Multivariate testing could shed light on the proportionate influence of relationships versus faith, when controlling for known factors that impact substance abuse treatment outcomes. Further testing could indicate which types of relationships and which kinds of relationship transactions (e.g., accountability, emotional support, spiritual mentoring) exert the greatest positive effect on outcomes. Secular substance abuse treatment programs of equivalent duration to GRMs should be included in effectiveness studies to assess the independent effect of relationships or other factors inherent in a longer treatment period when compared with faith-based treatment settings.

Limitations

The challenges of introducing a measurement and tracking system in a faith-saturated environment should not be underestimated. A planned national demonstration by NRI to test the effectiveness of faith-based substance abuse residential treatment will depend on consistent, thorough, and accurate data input by GRM staff. Improved protocols for data entry will decrease the attrition rate among the pilot’s sample, experienced as great as 50% or more on many pilot items. While the missing data speak almost entirely to provider variables rather than to participant characteristics that would bias findings, a great degree of statistical power is lost and the ability to analyze subgroups degraded by such sample loss.

While the NRI Pilot Study achieved its goals for measurement development and feasible data collection protocols, data quality suffered from understaffing and the tendency among many GRMs to employ their treatment programs graduates before persons with better qualifications. The provision of greater Web-based programming controls on data entry and error checking, and providing greater staff accountability and regular technical assistance via communication technology could improve sample retention throughout data collection.

Implications

Effectiveness findings in rigorously controlled trials of faith-based services may help clarify current public policy related to Charitable Choice, faith-based service delivery, or community-governed provision for need. Over time, with improved methods and more rigorous research, it is not unreasonable to expect a refinement of claims. As the knowledge base advances, faith-based service effectiveness within select populations, for some kinds of need, under particular service conditions, may be better understood.

161

ACKNOWLEDGMENTS

The National Recovery Initiative Pilot Project, an initiative of Atlanta Union Mission, Inc., is funded by two foundations: the National Christian Foundation with the recommendation of the Maclellan Foundation, and by a second corporate foundation that wishes their support to remain anonymous.

APPENDIX A: NATIONAL RECOVERY INITIATIVE PARTICIPANTS

Mission / Ministry Location

1. Fort Wayne Rescue Ministries Fort Wayne, IN

2. Trinity Mission Lafayette, IN

3. Phoenix Rescue Mission Phoenix, AZ

4. Water Street Rescue Mission Lancaster, PA

5. Wheeler Mission Ministries Indianapolis, IN

6. Miracle Hill Ministries Greenville, SC

7. Rescue Mission of Utica Utica, NY

8. Union Gospel Mission Twin Cities Minneapolis-St. Paul, MN

9. Memphis Union Mission Memphis, TN

10. Manna Inc. Bangor, ME

11. Light of Life Mission Pittsburgh, PA

12. Gospel Rescue Ministries Washington, DC

13. Charlotte Rescue Mission Charlotte, NC

14. Haven of Rest Akron, OH

15. Rescue Mission Alliance Ventura County, CA

16. Atlanta Union Mission Atlanta, GA

17. Wheeler Mission Ministries Indianapolis, IN

162

APPENDIX B: NATIONAL RECOVERY INITIATIVE (NRI) PILOT STUDY

SERVICES RECEIVED CHECKLIST ADDENDUM ©

Protocol: Below are definitions and examples of services to help you when filling out your checklist. Administration: This can be used with EITHER the Daily Services Received Checklist OR the Weekly Services Received Checklist. Missions may use either format based on what works best for them in order to ensure that participants fill out this information each day.

PERSON DISTRIBUTING FORM: It is EXTREMELY IMPORTANT that you read the following introduction word-for-word when you give the Services Received Checklist to a resident: This checklist is to help us know what services you receive as part of your recovery and how often you receive each service. It is extremely important that you complete a Services Received Checklist EVERY DAY that you participate in the recovery program. If you have any questions about any items on the Checklist, there is a Services Received Checklist ADDENDUM that explains and gives and example of everything on the Checklist.

1. Academic Services

Definition: Academic programs which may include but are not limited to reading, comprehension, vocabulary, spelling, memory, mathematics, writing, computer literacy and GED preparation that occur either in the agency or off-site

2. Life Skills Education

Definition: Various small groups or classes on life skills issues to help build and maintain appropriate relationships with self and others.

Examples: $ Personal Presentation $ Family Dynamics

$ Personal Financial accountability $ Inter-personal relationship skills

$ Nutrition $ Anger Management

$ Family Relationships $ HIV/AIDS education

$ Sexual Integrity Classes

3. Individual Counseling (by Mission staff)

Definition: One-on-one (minimum of 30 minutes including counseling participant documentation) between a participant and a program counselor regarding a significant recovery or life issue.

4. Mission-Related Supervised Work

Definition: Paid work role done by participant while residing at the Mission

163

5. Mission-Related Supervised Education/Training

Definition: classes towards an educational goal or for job skills done while residing at the Mission.

6. Relapse Prevention Education

Definition: Educational Classes or Groups led by qualified staff for Relapse Prevention

Examples:

F.A.S.T.E.R. Scale Worksheet

$ Forgetting Priorities

$ Anxiety

$ Speeding up

$ Ticked off

$ Exhausted

$ Relapse (total)

P.A.W.S. Education Class

$ Post

$ Acute

$ Withdrawal

$ Syndrome

7. Faith-Based or Conventional 12 Step Recovery Group

Definition: A twelve- step or comparable type of recovery group (internal or external to the agency)

Examples:

$ Alcoholics Victorious

$ Celebrate Recovery

$ Overcomers in Christ

$ Alcoholics Anonymous

$ Narcotics Anonymous

$ Sexual Addiction

8. Mentor Interaction

Definition: Telephone or face-to-face contact with a non-paid, not mission-staff confidant or helper who meets with a resident for any purpose on a regular basis

9. Resident Initiated Spiritual Activity

Definition: Any spiritual activity, broadly defined (e.g., prayer, Bible study, special topic study fostering spiritual growth and maturity) that occurs on a formal or informal basis as a result of resident(s) efforts. It may or may not include mission staff or volunteers as participants.

10. Mission-Organized/Supervised Recreation

Definition: Individual or group recreation to promote positive and enjoyable social interactions or physical recreation without substance abuse.

164

11. Mission-Organized/Supervised Physical Exercise

Definition: Physical Activity that works up a sweat.

12. Resident-Initiated Recreation

Definition: Individual or group recreation that is initiated by the resident to promote positive and enjoyable social interactions or physical recreation without substance abuse.

13. Resident-Initiated Physical Exercise

Definition: Physical Activity initiated by the individual resident that works up a sweat.

14. Individual Pastoral Counseling

Definition: Individual counseling with the mission chaplain or pastor in order for the client to develop a closer relationship with God.

15. Group Bible Study

Definition: Group of less than 15 that focuses on understanding, utilizing, and growing in the Word.

16. Chapel/Worship

Definition: Attendance at a chapel service or communal worship time at the mission.

17. Group Counseling (by Mission staff)

Definition: Group of less than 15 participating in interactive discussion lead by a qualified program counselor (excluding relapse prevention).

Examples of Group Counseling Topics:

$ Sexual Issues

$ Anger Issues

$ Chemical Dependency Treatment

$ Relationships/Community Living

$ Educational Progress

$ Discipleship Progress

$ Behavioral Issues

18. Employment Service

A. Pre-Employment Services

Definition: Services provided to clients prior to employment, which can include background checks, drug tests and assessments. These services allow employers to “check out” prospective employees before hiring them.

B. Employment Coaching

Definition: Provides tools and strategies to participants to assist in gaining or retaining employment. These strategies include implementing new skills, changes and actions to ensure participants achieve their targeted results.

Examples:

$ Resume writing

165

$ Interview Skills

$ Career Choices

$ Job Search Skills

19. Transportation

Definition: Providing a means of transport for clients to travel from one location to another

20. Drug Testing

Definition: A test to screen for drug and alcohol use performed randomly

21. Total Referrals and follow-through

Definition: Process outside of the Gospel Rescue Mission by which program counselor recommends services to the client to meet appropriate recovery needs AND the client follows through on the referral. Only count those referrals in which you know that the client responded and met with the referral source.

Examples:

$ Legal Services

$ Medical Services

$ Mental Health Services

$ Substance Abuse Services

$ After Care Services

$ Education Services

$ Peer-to-peer Recovery Support Services

$ Other _____________

APPENDIX C

Table C-1. Exposure to Faith-Related Program Elements, September 2007–February 2008

How many hours were each of the following services provided?

Service Mean SD

Individual Counseling (N=227) 7.17 9.20

Mission-Related Supervised Work 125.55 135.51

Faith-Based or Conventional 12-Step Recovery Group (N=269) 16.98 22.02

(continued)

166

Table C-1. Exposure to Faith-Related Program Elements, September 2007–February 2008 (continued)

How many hours were each of the following services provided?

Service Mean SD

Mentor Interaction (N=165) 12.47 23.66

Resident-Initiated Spiritual Activity (N=274) 25.23 40.69

Individual Pastoral Counseling (N=150) 4.34 5.77

Group Bible Study (N=322) 16.43 20.25

Chapel/Worship (N=347) 33.15 38.90

Group Counseling (N=235) 17.62 20.66

Figure C-1

Number of Faith-Based Services Recieved by GRM Participants

Cou

nt

120

100

80

60

40

20

0

Number of Faith-Based Services Recieved1-4 5-6 7 8-9

43 12.01%

125 34.92%

103 28.77%

167

87 24.3%

Figure C-2

Total amount of Faith-Based Services Used Across Selected Faith-Based Services by GRM Participants (units of service by sessions and hours)

Cou

nt

120

100

80

60

40

20

0

Amount of Faith-Based Service Units Received1-54 55-159 160-394 395+

89 25.14%

124 35.03%

83 23.45%

58 16.38%

Table C-2. Spiritual Growth Indicators in a GRM Faith Framework

How strongly do you agree or disagree with the following statements? N=348–353 (1=strongly disagree, 5=strongly agree) Mean SD

I need a change in my life. 4.81 0.53

I need Jesus to intervene in my life and provide guidance. 4.74 0.63

God is in control. 4.56 0.77

God is the authority. 4.73 0.63

God communicates His authority through Scripture. 4.63 0.69

I am willing to obey biblical principles. 4.32 0.78

I regularly apply biblical principles in my life. 3.59 0.98

168

Table C-3. Faith-Related Activities in a GRM Faith Framework

Please circle a number for each of the following questions, based on how often you engage in the following faith-related activities on a scale of 1 to 5. (1=never, 5=daily) In the last 30 days… Mean SD

How often did you attend worship services? N=330 2.81 1.31

How often did you participate in Christ-centered community, church, or small group? N=330 2.56 1.46

How often did you demonstrate Biblical stewardship of your money/income? (If currently without money, please circle 8) N=267 1.81 1.13

How often did you serve others on a voluntary basis? N=328 2.44 1.42

How often did you participate in a support group for the purpose of recovery? N=330 2.28 1.51

How often did you engage in group devotional activity or Bible study? N=330 2.45 1.50

How often did you engage in personal devotions or Bible study? N=331 2.83 1.57

How often did you participate in solitary prayer (exclusive of meals)? N=331 3.76 1.53

Table C-4. Personal Faith Qualities in a GRM Faith Framework

How often do you exhibit the following: N=328-332 (1=never, 5=very frequently) Mean SD

Acceptance of God’s forgiveness for my past actions 3.92 1.11

Acceptance of forgiveness from others 3.58 1.03

Forgiving others 3.88 0.92

Unselfishness/generosity 3.74 0.94

Helpfulness/servant attitude/kindness – (tenderhearted, sympathetic treatment to others) 4.00 0.86

Gratitude/Spirit of thankfulness 4.03 0.91

Offensive Language (reverse scored) 2.81 1.16

Respect for others/Non-judgmental attitude 4.01 0.87

Submission to authority 3.88 1.04

(continued)

169

Table C-4. Personal Faith Qualities in a GRM Faith Framework (continued)

How often do you exhibit the following: N=328-332 (1=never, 5=very frequently) Mean SD

Love – (Unconditional devotion to others) 3.91 0.97

Joy – (Feeling of happiness, delight, and great pleasure) 3.63 1.06

Peace – (Feeling secure, undisturbed, serene, and tranquil) 3.52 1.13

Patience – (Ability to wait or endure without complaint) 3.36 1.07

Faithfulness – (Consistent practice of what you believe and loyalty to people you value) 3.73 .93

Self-Control – (Ability to self-regulate and direct one’s actions and thoughts) 3.55 1.06

Table C-5. The Association between Relationship and Faith Measures among GRM Participants in SA Residential Treatment

Relationship Measure

Minimum Score

Maximum Score Mean

Statistically Significant Correlations with Faith Measures

Working Alliance Inventory, N=167 (high=positive)

134 252 226 Faith Maturity Score, r=.342, p=.000 Spiritual Growth Score, r=.193, p=016 Faith Activity Score, r=.417, p=.000 Personal Qualities Score, r=.298, p=000

Trust Staff Score, N=166 (high=positive)

1 20 15.11 Faith Maturity Score, r=.381, p=000 Spiritual Growth Score, r=.217, p=.006 Faith Activity Score, r=.007 Personal Qualities Score, r=.182, =.028

Lovejoy Life Skills Score, N=164 (high=positive)

8 20 14.12 Faith Maturity Score, r=.271, p=.001 Faith Activity Score, r=.266, p=.008 Personal Qualities Score, r=.365, p=.000

Rosenberg Self-Esteem Scale, N=163

11 22 16.09 Faith Maturity Score, r=.191, p=.018 Spiritual Growth Score, r=.018, p=.019 Personal Qualities Score, r=.201, p=.016

These low to moderate strength correlations show a consistently positive association with these measures of faith that is systematic and not due to chance.

170

APPENDIX D: TECHNICAL SUPPORTING DATA FOR MEASUREMENT DEVELOPMENT

Faith Measures

Faith Maturity Scale In the Pilot, 341 participants responded to the 16-item short version of the Faith Maturity Scale. The higher the score, the greater the faith maturity. Respondents show a low score of 20 and a high of 112, a mean of 77, and standard deviation of 21.58. As a published measure of long standing, it has acceptable psychometric properties.

Spiritual Growth Scale The seven items that constitute the newly developed Spiritual Growth Scale show excellent psychometric properties (Cronbach’s alpha=.82). The sum score of these items range from 10 to 35, with a mean score of 31.

Faith-Related Activities Scale This scale is also new and designed to fit the GRM faith framework for the faith practices aspect of individual change indicators for participants in SA residential treatment. These eight items also show good psychometric properties (Cronbach’s alpha=.87). The sum score of these items ranges from 8 to 40, with a mean of 20.

Personal Qualities Scale The 15 attitude and behavioral measures in Table 5 show high psychometric properties (Cronbach’s alpha=.92). The sum score of these items ranges from 15 to 70, with a mean of 53.

Relationship Measures

Trusting Staff Scale This measure uses five items to reflect the degree of trust placed in treatment staff (alpha=.77). The scores range from 1 to 20, with a mean of 15.11. These items include:

7. In the last 30 days… Nev

er

Les

s tha

n 1

time

per

mon

th

1-3

times

pe

r m

onth

At l

east

w

eekl

y

Dai

ly

How often were you able to talk to at least one staff member or volunteer openly and honestly about problems/issues of a personal nature?

1 2 3 4 5

How often did you feel safe (accepted by staff and volunteers) at this Gospel Rescue Mission? 1 2 3 4 5

(continued)

171

7. In the last 30 days… Nev

er

Les

s tha

n 1

time

per

mon

th

1-3

times

pe

r m

onth

At l

east

w

eekl

y

Dai

ly

How often did you seek out a staff member or volunteer to discuss a personal matter? 1 2 3 4 5

How often did you voluntarily share personal issues in a private or group meeting? 1 2 3 4 5

How often did you follow advice of a counselor/staff member or volunteer on a matter of personal importance? 1 2 3 4 5

Working Alliance Inventory The Working Alliance Inventory was completed by 167 respondents. It is a standardized, published scale on the staff relationship to a client most influential in the participant’s recovery. The WAI produced scores from 134 to 252 with a mean score of 226. The higher the score the stronger the relationship with this influential staff person. The scale items and format appear on the next page. As a published measure of long standing, it has acceptable psychometric properties.

Working Alliance Inventory (WAI)

Protocol: This instrument is to be self-reported by the resident.

(The remaining questions are to be answered only at 30 days, monthly thereafter, as well as at discharge, NOT AT INTAKE)

Below are 36 questions about your relationship with the staff person or Mission-related person that has most helped you to reach your goals. Remember that this information will be kept completely confidential. Your answers will not be connected with your name. The resident should NOT share this information with other staff or residents. 1. Which staff person or Mission-related person has most helped you to reach your goals? ___________________

2. What is that person’s role at the Mission? ___________________________________

The following questions all relate to your relationship with _________________ (that person)

Using the following scale rate the degree to which you agree with each statement, and record your answer in the space to the left of the item 1 = Not at all true

2 = A little true

3 = Slightly true

4 = Somewhat true

5 = Moderately true

6 = Considerably true

7 = Very true

___ I feel uncomfortable with ______.

___ ______ and I agree about the things I will need to do to help improve my situation.

172

___ I am worried about the outcome of these sessions.

___ What I am doing in our working relationship gives me new ways of looking at my problem.

___ ______ and I understand each other.

___ ______ perceives accurately what my goals are.

___ I find what I am doing in my work with ______ confusing.

___ I believe ______ likes me.

___ I wish ______ and I could clarify the purpose of our sessions.

___ I disagree with ______ about what I ought to get out of our sessions.

___ I believe the time ______ and I are spending together is not spent efficiently.

___ ______ does not understand what I am trying to accomplish.

___ I am clear on what my responsibilities are in our working relationship.

___ The goals of these sessions are important to me.

___ I find what ______ and I are doing are unrelated to my concerns.

___ I feel that the things I do in our working relationship will help me to accomplish the changes that I want.

___ I believe ______ is genuinely concerned for my welfare.

___ I am clear as to what ______ wants me to do in these sessions.

___ ______ and I respect each other.

___ I feel that ______ is not totally honest about his/her feelings toward me.

___ I am confident in ______’s ability to help me.

___ ______ and I are working towards mutually agreed upon goals.

___ I feel that ______ appreciates me.

___ We agree on what is important for me to work on.

___ As a result of these sessions I am clearer as to how I might be able to change.

___ ______ and I trust one another.

___ ______ and I have different ideas on what my problems are.

___ My relationship with ______ is very important to me.

___ I have the feeling that if I say or do the wrong things, ______ will stop working with me.

___ ______ and I collaborate on setting goals for my recovery.

___ I am frustrated by the things I am doing in our working relationship.

___ We have established a good understanding of the kind of changes that would be good for me.

___ The things that ______ is asking me to do don’t make sense.

___ I don’t know what to expect as the result of our working relationship.

___ I believe the way we are working with my problem is correct.

___ I feel that ______ cares about me even when I do things that he/she does not approve of.

173

Lovejoy’s Life Skills Scale How frequently do you do the following things? (Please circle your answer using the following scale: 0=Never, 1=Almost never, 2=Sometimes, 3=Fairly often, 4=Very often)

Openly express emotions with others

Have empathy(understanding) for the feelings of others

Reach out and ask for support when upset

Speak honestly

Life Skills Scale scores ranged from 8 to 20 with a mean of 14.11, with a sample size of 166. As a published measure of long standing, it has acceptable psychometric properties.

Rosenberg’s Self-Esteem Scale Please circle your answer using the following scale: 1=Strongly agree, 2=Agree, 3=Disagree, 4=Strongly disagree.

Strongly Agree Agree Disagree Strongly Disagree I feel that I am a person of worth, at least on an equal basis with others .......................................1 ......................... 2..........................3 ..........................4

I feel that I have a number of good qualities ...............1 ......................... 2..........................3 ..........................4

All in all, I am inclined to feel that I am a failure........................................................................1 ......................... 2..........................3 ..........................4

I am able to do things as well as most other people. ...............................................................1 ......................... 2..........................3 ..........................4

I take a positive attitude toward myself .......................1 ......................... 2..........................3 ..........................4

On the whole, I am satisfied with myself ....................1 ......................... 2..........................3 ..........................4

I wish I could have more respect for myself................1 ......................... 2..........................3 ..........................4

I certainly feel useless at times ....................................1 ......................... 2..........................3 ..........................4

At times I think I am no good at all .............................1 ......................... 2..........................3 ..........................4

Self-Esteem Scale scores ranged from 11 to 22 with a mean of 16.09, sample size of 163. As a published measure of long standing, it has acceptable psychometric properties.

174

NRI Trust Staff Scale Items (developed and tested with NRI Pilot data from GRMs)

N

ever

Les

s tha

n 1

time

per

mon

th

1-3

times

pe

r m

onth

At l

east

w

eekl

y

Dai

ly

How often were you able to talk to at least one staff member or volunteer openly and honestly about problems/issues of a personal nature?

1 2 3 4 5

How often did you feel safe(accepted by staff and volunteers) at this Gospel Rescue Mission? 1 2 3 4 5

How often did you seek out a staff member or volunteer to discuss a personal matter? 1 2 3 4 5

How often did you voluntarily share personal issues in a private or group meeting? 1 2 3 4 5

How often did you follow advice of a counselor/staff member or volunteer on a matter of personal importance? 1 2 3 4 5

REFERENCES

Ambrosino, R., Heffernan, J., Shuttlesworth, G., & Ambrosino, R. (Eds.). (2001). Social work and social welfare: An introduction. Belmont, CA: Wadsworth/Thompson Learning.

Benson, P. L., Donahue, M. J., & Erickson, J. A. (1993). The Faith Maturity Scale. Research in the Social Scientific Study of Religion. 5, 1–23.

Boddie, S.C., & Cnaan, R. A. (Eds.) (2006). Faith-based social services. Binghamton, NY: Haworth Press.

Chen, G. (2006). Social support, spiritual program, and addition recovery. International Journal of Offender Therapy and Comparative Criminology, 50(3), 306–323.

Cnaan, R. A., & Boddie, S. C. (2002). Charitable Choice and faith-based welfare: A call for social work. Social Work, 47(3), 224–235.

Dilulio, J. (2002). The three faith factors. The Public Interest, 129(4), 50–64.

Ebaugh, H. R., Chafetz, J. S., & Pipes, P. F. (2006). Where’s the faith in faith-based organizations? Measures and correlates of religiosity in faith-based social service coalitions. Social Forces, 84(4), 2259–2272.

Ellis, A., & Schoenfeld, E. (1990). Divine intervention and the treatment of chemical dependency. Journal of Substance Abuse, 2, 459-468.

175

Ellor, J. W., Netting, F. E., & Thibault, J. M. (1999). Understanding religious and spiritual aspects of human service practice. Columbia, SC: University of South Carolina Press.

Fowler, J. W. (1981). Stages of faith: The psychology of human development and the quest for meaning. New York: Harper & Row.

Gorsuch, R. L. (1995). Religious aspects of substance abuse and recovery. Journal of Social Issues, 51, 65–83.

Grettenberger, S. E., Bartkowski, J. P., & Smith, S. R. (2006). Evaluating the effectiveness of faith-based welfare agencies: Methodological challenges and possibilities. In S.C. Boddie & R.A. Cnaan (Eds.), Faith-based social services (p. 39–61). Binghamton, NY: Haworth Press.

Hertel, L. (1999) Man with a mission: Mel Trotter and his legacy for the rescue mission movement. Grand Rapids, MI: Kregel Publications.

Hodge, D. (2001). Spiritual assessment: A review of major qualitative methods and a new framework for assessing spirituality. Social Work, 46(3), 203–214.

Horvath, A. O., & Greenberg, L. S. (1989). Development and validation of the Working Alliance Inventory. Journal of Counseling Psychology, 36(2), 223–233.

Hugen, B., DeJong, F., & Venema, R. (2005). Faith makes a difference: A study of faith in human service programs. Acton Institute: Policy Form, Summer(6), 1–12.

Jeavons, T. H. (1994). When the bottom line is faithfulness. Bloomington, IN: Indiana University Press.

Jeavons, T. H. (1998). Identifying Characteristics of religious organizations: An exploratory proposal. In N.J. Demerath III et al. (Eds.), Sacred companies: Organizational aspects of religion and religious organizations (pp. 79–95). New York: Oxford University Press.

Johnson, B. R. (2002). Objective hope assessing the effectiveness of faith based organizations: A review of the literature. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

Karger, H. J., & D. Stoesz (2005). American social welfare policy: A pluralist approach. 4th ed. Boston: Pearson Educational.

Koenig, H. G. (Ed.). (1998). Handbook of religion and mental health. San Diego, CA: Academic Press.

Koenig, H. G., McCullough, M. E., & Larson, D. B. (2001). Handbook of religion and health. New York: Oxford University Press.

Lovejoy, M. (1995). Patients’ perspective on the process of change in substance abuse treatment. Journal of Substance Abuse Treatment, 12, 69–82.

176

Meacham, J. (2006). American gospel: God, the founding fathers, and the making of a nation. New York: Random House.

Miller, W. R. (1997). Spiritual aspects of addiction treatment and research. Mind/Body Medicine, 2, 37–43.

Neff, J. A., & MacMaster, S. A. (2005). Spiritual mechanisms underlying substance abuse behavior change in faith-based substance abuse treatment. Journal of Social Work Practice in Addictions, 5(3), 33–54.

Neff, J. A., Shorkey, C. T., & Windsor, L. C. (2006). Contrasting faith-based and traditional substance abuse treatment programs. Journal of Substance Abuse Treatment, 30(1), 49–60.

Pardini, D. A., Plante, T. G., Sherman, A., & Stump, J. E. (2000). Religious faith and spirituality in substance abuse recovery: Determining the mental health benefits. Journal of Substance Abuse Treatment 19, 347–354.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton; University Press.

Sider, R. J., & Unruh, H. R. (2004). Typology of religious characteristics of social service and educational organizations and programs. Nonprofit and Voluntary Sector Quarterly, 33(1), 109–134.

Smith, S. R., Sosin, M. R. (2001). The varieties of faith-related agencies. Public Administration Review, 61(6), 651–670.

Specht, H., & Courtney, M. (1994). Unfaithful angels: How social work has abandoned its mission. New York: Free Press.

Unruh, H. R., & Sider, R. (2005). Saving souls, serving society. New York: Oxford University Press.

von Furstenberg, G.M. (2006). Detecting and decomposing the faith factor in social-service provision and absorption. In S.C. Boddie & R.A. Cnaan (Eds.), Faith-based services: Measures, assessments, and effectiveness. Binghamton, NY: Haworth Press.

Warfield, R. D., & Goldstein, M. B. (1996). Spirituality: The key to recovery from alcoholism. Counseling and Values, 40, 196–205.

Zannis, D. A. & Canaan, R. A. (2006). Social service research and religion: Thoughts about how to measure intervention-based impact. In S.C. Boddie & R.A. Cnaan (Eds.), Faith-based services: Measures, assessments, and effectiveness, Binghamton, NY: Haworth Press.

177

IN GOD WE TRUST, ALL OTHERS BRING DATA: ASSESSING THE STATE OF OUTCOMES MEASUREMENT

FOR FAITH-BASED AND COMMUNITY-BASED PROGRAMMING

Robert L. Fischer

After being set in motion by the Charitable Choice provision of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, and expanded and institutionalized by executive orders beginning in 2001, the Faith-Based and Community Initiative (FBCI) has sought to encourage greater participation by faith-based and community organizations (FBCOs) in federally funded human service programming. Since 2001, the FBCI has aimed to give these organizations equal opportunity with secular and larger organizations to secure federal funding for the delivery of social services. The expanded role for FBCOs within the domain of federally funded human services has carried with it an increased interest in the capacity and operational effectiveness of these organizations.

During this period, the existing literature on FBCO programming and research on all aspects of faith in the provision of human services has grown dramatically. One portion of the growing evidence base pertains to the effectiveness of the services delivered by FBCOs. Approaching the task of understanding outcomes in the FBCO arena is not without marked challenge. In their book on FBCO welfare-to-work programs, Monsma and Soper (2006) concluded in a chapter titled “The Effectiveness Muddle” that “effectiveness—even when defined in terms of program outcomes—is a complex concept that is often misunderstood. There are methodological problems in operationalizing a study that seeks to measure program outcomes from different types of programs in order to compare their relative effectiveness. There are few existing theories or studies to guide us. Nevertheless, public policy makers constantly seek to make decisions based on the presumed relative effectiveness of different types of human services programs” (p. 37).

Given the relative youth of the FBCI, and the reality that the development of the capacity of FBCO programs will take time, the growth of the research literature should be seen as evolutionary. In the context of the limited state of current research, important questions need to be addressed about how FBCOs conduct their programs and the extent to which the programs achieve the intended outcomes. The overarching interest of this paper is in programs that have been the emphasis of the federal FBCI, both faith-based and secular. However, this review

179

concentrates nearly exclusively on faith-based programs, as these programs have been the subject of the most focused study. This study seeks to assess the current evidence base by examining the previous relevant reviews, highlighting noteworthy studies in the field, and conducting a synthesis of the available research on FBCO effectiveness. The analysis places a premium on examining the available evidence for the purpose of effectively informing policy and practice as it relates to the FBCI.

CONTEXT FOR RESEARCH ON THE EFFECTIVENESS OF FAITH-BASED AND COMMUNITY-BASED SERVICES

The FBCI is rooted in the notion of drawing on the natural capacities and strengths of FBCOs to deliver effective programming. These organizations, as indigenous entities with staff who often reside in the surrounding neighborhoods, are seen as having invaluable connections and credibility within their communities. FBCOs often have “direct and consistent contact” with the most needy in their communities and their articulated mission serves to dedicate their efforts to serving these needs as best they can (Fink & Branch, 2005, p. 1). Also, FBCOs have established roots and connections both within the geographic area where they are located as well as within the broader faith communities.

FBCOs are recognized as having particular expertise and advantage in some areas. For example, in its narrative on the Compassion Capital Fund, the U.S. Department of Health and Human Services (DHSS) describes FBCOs as being “uniquely situated” to serve “families in poverty, prisoners reentering the community and their families, children of prisoners, homeless families, and at-risk youth” (DHHS, 2002). Beyond simply acknowledging expertise, however, some proponents outside of the FBCI have gone further, arguing that faith-based programs are in fact superior to the conventional alternatives in terms of effectiveness and cost (Cnaan & Boddie, 2006; Smith, Bartkowski, & Grettenberger, 2005). These claims appear to be based on the experiences of single studies or compelling anecdotes, rather than on a systematic review of the evidence. The dialogue about the prospective effectiveness of FBCO programming has had a tendency to outpace the actual data available, or to focus on effects for subgroups of participants (Singer & Friel, 2007; Wall Street Journal, 2003). This situation underscores the need for additional research on the outcomes of FBCO-sponsored programs, and specifically in comparison to conventional programming or the absence of programming entirely.

In particular, the central question of the unique effectiveness of faith-based organizations in providing human services is also confounded by the role that religion or faith may play in the lives of individuals. Simply examining the effects of programs that have faith as an element of

180

their delivery will not shed light on the central question. The issue is far more circumscribed (see Figure A-1 in Appendix A.) All individuals possess personal characteristics that pertain to their level of faith and their engagement in religious practice. The existing research shows that the association between religious engagement and positive behaviors and well-being is fairly clear— the relationship is significant and positive but not conclusively causative (e.g., Johnson, Tompkins, & Webb, 2002). Although religious engagement is associated with indicators of better health and behaviors, it is also plausible that the individuals who choose healthy behaviors also tend to choose religious engagement (hence the multidirectional arrows in Figure A-1). When the notion of an FBCO program is introduced into these relationships, it becomes apparent that the individual’s choice about what services to use may reflect something about their faith disposition. Although the FBCO program may be influencing the behaviors and well-being of the individual, these changes may also be related to the underlying characteristics of the individuals served.

The ability to distinguish between the effect of this so-called selection bias and true treatment effect depends heavily on the research methodology that is undertaken (Berk, 1983). In a study that specifically examined the factors associated with individuals volunteering for a faith-based program, the authors suggest that researchers can only effectively address this issue through more rigorous study designs involving a sound comparison between served and nonserved groups (Camp, Klein-Saffran, Kwon, Daggett, & Joseph, 2006). In addition, beyond the characteristics of the organization there are also the characteristics of the individual program. For example, an FBCO may deliver different programs that possess different levels of faith intensity, and the faith dimension at both these levels can play a role in participant well-being. Defining or categorizing programs by the presence of faith is the subject of ongoing debate (Ebaugh, Pipes, Chafetz, & Daniels, 2003; Smith & Sosin, 2001; Twombly, 2002).

Further, the particular interest in the context of the FBCI is whether the programs that have been the emphasis of the initiative thus far are effective. The FBCI was introduced in part to “level the playing field” in the competition for federal funding, acknowledging that historically many FBCOs had been unable to access these funds. Some FBCOs, such as Lutheran Family Services, Catholic Charities, and Jewish Family Services, found ways to successfully compete for federal funds. These large organizations had developed the internal capacity to deliver and manage services in a manner that satisfied federal funding restrictions. Thus, the focal interest now is on the effectiveness of programs that were not well represented in the previous federal funding picture.

181

This paper focuses on two key questions. First, to what extent are FBCO services effective? For the purposes here, effectiveness is defined as a program’s ability to have participants show progress in achieving the desired outcomes (i.e., changes in knowledge, attitude, behavior, or status). For example, do children served by a mentoring program delivered by an FBCO show fewer behavioral problems and/or improved academic performance during or after the mentoring experience? The types of studies that are relevant to address this question are those that collect data on participants using some sort of outcomes measurement approach, which is often limited to pre-post assessments on a single group of participants (Bartowski, Call, Heaton, & Forste, 2007; Hangley & McClanahan, 2002; MacMaster et al., 2007; Rock, 2002; Roman, Wolff, Correa, & Buck, 2007). Although useful, these studies routinely lack outcome data from a comparison group that would provide a benchmark of what would have happened in the absence of the program. As such, these studies can only indicate whether participants changed during the time frame of the intervention and cannot address whether the change was a result of the program’s efforts.

The second question addresses to what extent are FBCO services more effective than other approaches? For example, do clients served by different types of welfare-to-work programs (e.g., faith-based versus secular) show different levels of success in achieving the desired employment outcomes? The types of evidence that will inform this question are based on more rigorous evaluation designs in which meaningful comparisons can be made to a group of comparable individuals who received alternate services or usual care. This requirement is much more restrictive, and the number of studies that will achieve this standard is necessarily fewer than those that will address the first question. The most conclusive designs involve random assignment to groups or other quasi-experimental approaches to creating equivalent groups.

TAKING STOCK OF WHAT IS KNOWN ABOUT FAITH-BASED AND COMMUNITY-BASED SERVICES

The present review comprises three steps: (1) a description of the previous reviews that have examined research in the FBCO domain, (2) a review of several study exemplars that illustrate particular strengths present in the literature, and (3) a review and synthesis of the existing comparative studies on FBCO effectiveness.

Learning from Previous Reviews

Research reviews are an important building block in developing a research literature, in that they offer a periodic reflection on the state of the evidence base. In the FBCO arena, four previous reviews were identified and each is very briefly summarized in Table B-1 (see Appendix B).

182

Each of these focused on reviewing the characteristics and findings of research on FBCO services. These prior efforts sought to review the existing research base, describe its status, and summarize its contents narratively. None of the efforts attempted to quantitatively summarize evidence on program effects.

A critical and influential first review, entitled “Objective Hope,” was done by Johnson, Tompkins, and Webb in 2002. The report presents a review of nearly 800 studies, including a core group of 25 studies examining the effectiveness of FBCO services. The remainder of these studies related either to the relationship between religion and health outcomes or to the relationship between religion and other forms of well-being. Based on the quantity and quality of literature existing at the time on FBCO effectiveness, the authors concluded that although the overall body of work showed promising effects, most areas of FBCO service “have not been the subject of serious evaluation research” (p.21).

In 2003, the Roundtable on Religion and Social Policy released a review that focused on literature related to the scope and scale of faith-based and community-based activities (Scott, Montiel, Keyes-Williams, & Han, 2003). Although the primary intent was to summarize efforts at distinguishing between different types of FBCOs, describe their activities, and explore the amount of resources FBCOs were contributing through their work, the review included studies that also examined FBCO effectiveness. The authors grouped studies according to whether they focused on a specific locale or region, a multistate area, or a national perspective. The review included studies of FBCOs as well as congregations. The authors concluded that additional work was needed to document the specific coverage of FBCOs and to develop more complete data on the scope and scale of FBCO activities and comparative data from other service providers in a given geography. This review was focused on detecting the relative range of FBCO programming in the social service landscape, rather than on investigating the effectiveness of those programs.

In 2004, an additional review was published that specifically focused on studies of health programs in faith-based organizations and their effectiveness (DeHaven, Hunter, Wilder, Walton, & Berry, 2004). The types of health programs included in the review span the topics of general health, cardiovascular health, cancer, mental health, and nutrition. Approximately 43% of the programs involved the provision of health education or services in a church setting and an additional 25% were programs emerging from a congregation’s health ministry; the remaining 32% combined faith-based and secular elements. The authors reported that outcomes were measured in 28 (53%) of the studies and for each they reported whether the results were statistically significant. The authors concluded that faith-based health programs “can produce

183

positive effects” and reported that among the 16 studies that reported a statistical test, 15 showed a statistically significant difference favoring the intervention group compared with unserved populations (on outcomes such as decreased cholesterol and blood pressure and increased health and fruit/vegetable consumption). The authors also recommended strategies for increasing the number of effectiveness studies and their usefulness in health programming with diverse communities.

The most recent review, conducted by Ferguson, Wu, Spruijt-Metz, & Dyrness (2007), used the systematic review approach to examine how effectiveness has been defined in research on FBCOs. Using the keywords “faith-based” and “program effectiveness,” the authors searched a range of electronic databases as well as the Web sites of known institutional sponsors and producers of such research. The search ultimately identified 29 studies: 21 studies had a quantitative component and 8 studies used a comparative design. The authors discuss how effectiveness has been defined across these studies and offer a narrative summary of the findings within each of six outcome areas (e.g., health, criminal recidivism). They ultimately concluded, based on the limited number and quality of prior studies, that “the quality of findings from some previous evaluation studies on the effectiveness of faith-based programs remains questionable” (p. 272). They offered a number of useful recommendations for the field, such as working to broaden outcomes beyond the client level and more clearly describing the role of faith in program models.

Collectively, these four reviews reflect the state of the literature at the time of their completion as well as the needs for additional research. Combined, the reviews suggest three general observations. First, engagement in religious behaviors is convincingly associated with numerous indicators of positive health and well-being. The majority of these studies are correlational in nature; thus, they do not control for other factors. Second, there is a growing body of evidence that participants in FBCO programming do show improvement on identified outcomes over the course of their involvement with these programs. The limited number of comparative studies shows that FBCO-served populations may fare better in relation to comparison groups in some aspects. Third, the prior two observations do not provide sufficient evidence for documenting the benefit of FBCO programming. Rather, comparative studies using well-constructed reference groups are needed to more fully illuminate the effectiveness dimension.

Examining Study Exemplars

Despite the limitations of the literature in this area, there are a number of noteworthy studies that have contributed considerably to the field of FBCO effectiveness research. The criteria for

184

identifying these exemplars were subjective; however, each study was selected to illustrate a particular contribution to the current knowledge base on FBCOs. The distinct contribution of each study will be described here. Three of these studies have been cited extensively in the literature for their findings and for the research methods they used. These three studies were also included in the review conducted by Ferguson et al. (2007). The fourth study highlighted was completed more recently in 2007 and is included because it used random assignment to groups.

The first study of note is Monsma & Soper’s (2003, 2006) comparative evaluation of a range of welfare-to-work programs operated in Los Angeles. This study focused on comparing how welfare recipients served by programs operated by various types of providers differed in terms of their characteristics and their labor force outcomes. The authors selected 17 employment programs and then collected participant-level data from the selected programs. Although based on an observational design subject to selection bias concerns, this study offers a systematic and comprehensive look at a set of service providers all working toward the same set of outcomes, including two categories of faith-based employment programs.

Based on this analysis, the authors concluded that participants in the for-profit programs generally had fewer barriers to employment (e.g., criminal history, substance abuse, mental health issue), whereas participants in the programs with the more intensive faith elements generally had more barriers. The study found that the faith-based programs were the least successful in placing unemployed participants into jobs by 6 and 12 months, but were somewhat more effective in helping employed participants retain their employment once placed.

A second study is the evaluation of Amachi, a model mentoring program for children of prisoners in Philadelphia. The Amachi program is based on a partnership between secular and faith-based organizations (Jucovy, 2003). This study focused on examining how the model was implemented and how the mentoring relationships fared over time. Amachi adopted the outcome model used by Big Brothers–Big Sisters (BBBS), which previously found that positive results for mentees begin to occur after 12 months of engagement. Using this benchmark, 62% of Amachi matches were active 12 months or longer and exceeded engagement in the generic BBBS program (46%). This study demonstrated how faith-based services can adopt secular outcomes measurement approaches from the same program domain.

A third notable study involves the InnerChange Freedom Initiative (IFI), a faith-based prerelease program for prisoners operated in Texas (Johnson & Larson, 2003). The intent of this study was to assess the implementation of the program and measure participant outcomes. The program involves in-prison, Bible-based programming and 6 to 12 months of aftercare once the

185

participant is paroled. The authors implemented a matched group design, whereby outcomes for IFI participants (and graduates) could be contrasted with three comparison groups of prisoners who were similar to the IFI group in important ways.

Overall, the study found that IFI participants showed equivalent or slightly more recidivism than the three comparison groups on measures of rearrest and reincarceration within 2 years of release. However, the authors reported that program graduates demonstrated dramatically lower recidivism rates over 2 years compared with noncompleters—17% versus 50% on rearrest and 8% versus 36% on reincarceration, respectively (Johnson & Larson, 2003). Although the experience of graduates does not generalize to all program participants, the contrast with noncompleters may be important from a programmatic stance. Examining the characteristics of those who complete may suggest whom the program engages most successfully. For example, in the case of IFI, program graduation was more frequent among Hispanic men, men over 35 years of age, and men judged to be at low risk, compared with the other categories of participants.

A fourth exemplary study involved a faith-based intervention with cocaine-addicted women (Stahler, Kirby, & Kerwin, 2007). Although based on a very small sample size of 18, the study showed how random assignment could be used effectively to compare treatment alternatives. Following a recruitment and screening procedure, the 18 study volunteers were randomly assigned to the two groups. Both groups of women received the supervised residential treatment component of the program. In addition, the intervention group received faith-based workshops and mentoring (called the Bridges program), while the control group was exposed to other group programming of a secular nature. The study showed significantly better program retention among the Bridges group, as well as higher rates of drug abstinence at 6 months postintake based on urine samples (75% versus 30%). However, the differences between the groups were not statistically significant at 3 months postintake.

Summarizing the Available Studies

A final stage of this work is to engage in a synthesis of the existing comparative research. The task of reviewing the available evidence on the effectiveness of FBCOs is a process of both inclusion and exclusion. To clarify what is known on this topic, decisions were made as to what evidence is relevant and what studies can inform the process. Three categories of research were not deemed sufficiently relevant to the present policy discussion and were not reviewed as part of this effort. The three domains—studies of religiosity, studies of programs targeting church congregations, and studies that are descriptive in nature or only use qualitative methods—are

186

described in Appendix C. These study domains are important, but they are less relevant to answering more narrow questions about program effectiveness.

METHODS

The study methods used here are based on the accepted standards for research synthesis approaches (Lipsey & Wilson, 2001). The framework of the research synthesis procedure can be cast in terms of six interrelated steps: (1) specification of the questions of interest, (2) specification of the research domain, (3) retrieval of empirical studies, (4) review and coding of studies, (5) statistical aggregation of research results, and (6) presentation and interpretation of findings (Cordray & Fischer, 1994).

With the primary focus of the review effort delineated, a search was conducted to locate as many candidate studies as possible for inclusion in the review. The search had two main components. First, the existing research reviews were examined and the studies included in each of these were culled for candidates for the present review (DeHaven et al., 2004; Ferguson et al., 2007; Johnson et al., 2002; Scott et al., 2003). The availability of prior reviews is a major asset and these offer a key starting point for the search procedure.

Second, a search was conducted of 10 large electronic databases containing publications and reports in the social and behavioral sciences.1 The disciplines represented in these databases include anthropology, economics, law, medicine, nursing, policy studies, psychology, social work, and sociology. The search was focused using the search terms “faith-based,” “community,” and “evaluation” together, as well as “faith-based” and “outcome” together. Studies targeted for inclusion had all of the following attributes: (1) reported on the evaluation of a program delivered by a faith-based or community-based organization,2 (2) reported quantitative data on participant outcomes relevant to the intervention (e.g., employment, substance abuse, criminal activity), (3) used a comparative study design involving a group of program participants and a group of individuals who participated in an alternate program or no special programming, (4) involved the evaluation of a human service programming effort similar in substantive nature to those targeted by the FBCI, (5) involved programming that was directed to a general population of individuals rather than a group composed of members of a religious congregation, and (6) was reported in 1990 or later. Of particular interest were studies completed after 2004 because the existing reviews by Johnson et al. (2002) and Ferguson et al. (2007) had identified the vast majority of potential studies completed through 2004. The focus of this synthesis is on studies that examine the outcomes of faith-based programs, often in comparison to secular programs with the same objectives.

187

The present review differs from the prior review efforts in two important ways. First, this review specifically targets only those studies that made use of a comparative design in assessing quantitative outcomes. The goal here was to restrict attention to the potentially most rigorous studies that could address the issue of effectiveness. For example, a group of studies included in the review by Ferguson et al. (2007) were excluded here, including eight studies that used only qualitative methods and an additional four studies that relied on a single-group design. An additional 10 single-group studies in the review by Johnson et al. (2002) were also excluded. Second, the present review extends the search window by three additional years to 2007. This additional period is quite important, especially in a field that has seen such dramatic growth and attention over the past few years.

Based on the search procedures, a total of 92 independent citations were generated. Once identified, copies of publications and reports were secured and examined for eligibility by the author. In this process, particular emphasis was given to the type of research design used in the study, distinguishing between studies that used a single-group pre-post or related design and those that were truly comparative in nature (i.e., using two or more groups). A supplemental search technique was also applied by reviewing the citations and footnotes of all retrieved studies. Studies that were ultimately deemed to meet the inclusion criteria were then formally reviewed.

The review of the existing literature identified a meaningful core of studies that examined the outcomes of FBCO programming. However, only 18 of the 92 quantitative outcomes studies identified used a comparative research design. These studies are summarized in Table D-1 (see Appendix D). Among the 18 studies identified, 13 had been included in at least one of the prior reviews. For each study, the table highlights the target population, the study design and sample size, a brief statement of the intervention, the outcomes measures, the basic findings, whether subgroup analyses were conducted, and any relevant effect size information. The 18 studies span six distinct target populations, with multiple studies focused on prisoners and former prisoners (8), welfare clients (4), substance abusers (2), the elderly (2), and additional studies of children of prisoners and Latino women. Sample sizes vary dramatically, as do the procedures used to construct a comparison group for the studies. The key outcomes of interest are specific to the substantive focus of the programs. Although there is some consistency within program domains (e.g., recidivism among prisoner programs), the time frames of the outcomes vary. Finally, subgroup analyses were examined in over two thirds of the studies, but these varied in scope and focus.

188

For each of the studies, the reported outcomes data were collected and used to compute standardized mean effects on the key program outcomes identified by the study’s authors. The effect size provides a standardized metric for comparing across studies and outcomes. Of the 18 studies, 14 compared faith-based programming of some type with either a specific secular program or generally available services delivered to the same target population. The remaining four studies compared a faith-based program with the experience of individuals served by no specific program.

The unweighted mean effect size of all 18 studies is .216 and ranges from a high of .736 to a low of −.807. Effect sizes with a positive value indicate that the intervention group had greater success on the outcomes, while negative values show better success in the comparison group. The unweighted effect size does not take into account the differing sample sizes underlying the studies, as studies based on larger samples tend to provide more precise estimates and should be given more weight in this approach. When weighted for the study sample size (and after removing one study deemed to be a statistical outlier), the mean effect size is .205. Figure E-1 (see Appendix E) presents a summary of distribution of effect sizes. The overall mean effect size would be judged in the small range according to the standard offered by Cohen (1988) and in the bottom 25% of effects according to Lipsey and Wilson (2001). Another interpretation of an effect size of .2 is that it equates to a 10% difference on a success measure between the treatment and comparison groups (e.g., 55% versus 45%).

Two of the larger subgroups of studies were also examined. The eight studies of interventions with prisoners and former prisoners yielded a weighted mean effect size of .138, again in the small range. The three studies of welfare-to-work programming (after removing the one statistical outlier) yielded a weighted mean effect size of .503. This effect is in the moderate range and is equivalent to a 24% difference in the success rates between groups (e.g., 62% versus 38%).

The set of available studies that use a comparative design and report quantitative outcomes in sufficient detail is relatively small and is only able to support a preliminary quantitative synthesis of findings. As such, at present, the findings presented should be seen as illustrative of the kinds of beneficial impacts that FBCOs can produce across a range of substantive domains working with differing target populations. On average, the presence of faith in these programs, along with the other characteristics of the interventions, appears to result in a modest effect on the outcomes of interest over those observed among comparison populations.

189

FINDINGS AND IMPLICATIONS

Since its launch in 2001, the FBCI has dramatically increased the role of smaller FBCOs in the delivery of federally funded social services (White House, 2008). Concurrently, there has been expanded interest in the ability of FBCOs to document their outcomes, and in identifying high-quality research to test the effectiveness of services delivered by FBCOs. Despite considerable effort and progress, the existing evidence base remains limited. A preliminary quantitative synthesis shows that the overall effect of FBCO programs, although modest in size, demonstrates that these programs tend to produce somewhat better outcomes compared with usual services, secular services, or no special programming. More data are needed to confirm this finding, as are data to investigate the effectiveness of specific categories of programs for defined target populations. In order to further contribute to existing knowledge in this domain, efforts should be continued in at least three areas, as described below.

Outcomes Measurement and FBCO Capacity

Several challenges to evaluating the effectiveness of FBCO services stem from the fact that FBCOs targeted since 2001 tend to be small nonprofits with limited capacity for ongoing data collection or systematic research studies (Fischer, 2004). The growth of outcomes measurement has spurred a major shift in the way nonprofits view their work and the way they communicate their work to their funders, clients, and other stakeholders (Fischer, 2001; Hatry, Van Houten, Plantz, & Greenway, 1996). The Compassion Capital Fund National Resource Center (2005) has recognized the value of this approach and has produced a manual on outcomes measurement for use by intermediary organizations assisting FBCOs to build capacity.

The available evidence suggests that because of their limited size and relative inexperience with outcomes measurement, many FBCOs need specific assistance to develop capacity to collect, manage, and analyze their data. Johnson et al. (2002) reported that FBCO supporters often cite exceptionally high rates of success for programs, but that “closer examination of these accounts…tends to reveal mere simple summary statistics based on in-house data compiled by the religious organizations and ministries themselves” (p. 15). FBCOs should strive to collect more complete and accurate data and present it in a methodologically rigorous and neutral fashion. Some researchers have called for federal policy makers to emphasize the need for more and better evaluation of FBCO services, going so far as to suggest a requirement for clear logic models along with financial support for evaluation efforts (Mears, Roman, Wolff, & Buck, 2006).

190

A central implication of the capacity issue is the imperative of addressing the developmental needs of FBCOs. The recognition that FBCOs require specialized assistance in fully developing and assessing their programs has resulted in the funding of intermediary organizations to help build the capacity of FBCOs (Sherman, 2002). For example, of the Compassionate Capital Fund monies initially appropriated, $25 million (83%) was for intermediary organizations to aid FBCOs “to replicate or expand best practices and model programs in targeted areas” (Sherman, 2002). As the capacity of FBCOs is better understood, there can be better planning to address their capacity-related needs (Clerkin & Gronbjerg, 2007; Leake et al., 2007).

There is a strong emphasis on working to increase the capacity of FBCOs through promoting internal development and external support via intermediary organizations (Fink & Branch, 2005; Sherman, 2006). In fact, the strategy is now recognized as one of the key innovations of the FBCI (White House, 2008). Because organizational capacity is inextricably linked to an organization’s ability to document its outcomes and take part in more rigorous research, investments in FBCO capacity will facilitate further development of the research literature as well.

Outcomes measurement may also prove to be an empowerment mechanism for FBCOs. Fagan, Horn, Edwards, Woods, and Caprara (2007) suggest that “outcome-based evaluation has the potential to engender a revolution of increased effectiveness in the faith community and debunk skeptics’ claim that faith-based programs are only about ‘feel good’ results rather than producing solid and measurable impacts” (p. 1).

Rigorous Evaluation

The ongoing dialogue over the FBCI involves a range of concerns but continues to include a heavy emphasis on the effectiveness of FBCO services compared with both their secular counterparts and/or conventional services (DiIulio, 2002). The most promising avenues for responding to the data needs regarding FBCOs are through improving and expanding data collection practices and fielding more rigorous comparative studies to address issues of effectiveness. In the evaluation of any human service intervention, it is widely accepted that the use of randomized, well-implemented research designs leads to the most credible assessments of program impact. Experimental and quasi-experimental research designs have the distinct advantage of eliminating the role of a range of plausible intervening factors that could compete with the program in explaining impacts (Cook & Campbell, 1979). To date, only one study of faith-based services using a randomized design has emerged (Stahler et al., 2007).

191

Federally funded research on FBCI-related programming may hold the most promise in regard to expanding the use of more rigorous designs. As highlighted in a recent White House report in a chapter called “Measurement Matters,” no less than 20 studies of FBCI-related programs are currently underway by nine federal agencies (White House, 2008). These should be monitored to gauge what they may be revealing about effective programming strategies. The evaluation research community and funders of such efforts should expand the discussion of how to bring more resources, both technical and financial, to bear on the evaluation needs of the FBCO sector. These efforts will necessarily involve governmental and contractual oversight by funders to monitor the execution of FBCI activities, but must also get at the intended participant outcomes (Government Accountability Office, 2006). In addition, although more rigorous research is needed, this should not preclude further important advancements based on qualitative and observational research approaches. These research domains help bring clarity to the context of these programs and the lives of participants, and offer unique and richer understanding of program delivery and effectiveness.

Operationalizing Faith

An area of great interest and debate has been in characterizing the nature of faith-based programs (e.g., Monsma & Mounts, 2002; Sider & Unruh, 2004; Smith & Sosin, 2001; Working Group on Human Needs, 2002). For example, the Working Group (2003) defines an FBCO as “any entity that is self-identified as motivated by or founded on religious conviction” (p.2). The ability to assess the relative degree of faith intensity of a social service program is central to clarifying the program’s theory, logic, and ultimately the key outcomes. If the role of faith is a key ingredient in the expected success of the faith-based programs, then it is essential to better understand and measure its presence (Fischer, 2004). Faith can be both a matter of the context or environment of programs as well as part of the intervention itself, and as yet there are very limited data on this distinction.

CONCLUSION

Despite the substantial growth in the funding available to faith-based and community-based organizations over the past decade, the field of research on FBCOs remains very young and underdeveloped. As recently as 2002, the U.S. General Accounting Office concluded that the literature “provides no information on which to assess the effectiveness of FBCOs as providers of social service” (p. 17). Although advances have been made and a productive dialogue is underway, the extent of the existing evidence base is insufficient as a guide for program planning and enhancement. Overall, because of the relative youth of the FBCO research field, there is a

192

lack of systematic data on FBCO services and their effectiveness. As indicated earlier, much of the existing research on FBCO services is descriptive in nature, with a focus on programmatic models, delivery styles, and funding streams of FBCO services (Independent Sector, 2003). However, the research in the field has expanded over time (1998 to 2007) and now includes a number of key empirical studies as well as important efforts to distill what is known about the effectiveness of existing programs. The field needs to move forward with an agenda of establishing data systems for the purposes of accountability, program improvement, and demonstrating effectiveness.

ACKNOWLEDGMENTS

The author wishes to thank Grace Dyrness, Kristin Ferguson, and Mark Joseph for helpful comments provided on an earlier draft, as well as Diwakar Vadapalli and Monica Craven for assistance in preparing the manuscript.

NOTES

1. The databases included PsychINFO, Social Science Index, Social Work Abstracts, BiblioLine, Lexis-Nexis, MEDLINE, Dissertation and Theses Abstracts, Nursing and Allied Health Sources, Health Source, and Arts and Humanities.

2. The focus here is on what Chaves (2002) calls “religious nonprofits,” defined as “religious organizations working in nonreligious functional fields includ[ing] the wide variety of religious organizations doing virtually everything secular nonprofit organizations do” (p.1524).

193

APPENDIX A

Figure A-1. Linkages Between Faith and Wellness

Religious

Engagement

Faith-Based and

Community Intervention

Individual

Faith Intensity

Positive Behavioral

Outcomes and Well-being

194

APPENDIX B

Table B-1. Reviews of Research Related to the Faith-Based and Community Initiative

Authors &Title Year

Number of all

studies reviewed

Number of quantitative

studies included

Date range of studies included Focus of studies included

Johnson, Tompkins, & Webb—Objective Hope

2002 766 25 1933–2002 Relationship between religion and health and well-being outcomes; efficacy of FBCOs

Scott, Montiel, Keyes-Williams, & Han—The Scope and Scale of Faith-Based Social Services

2003 75 Not a specific focus

1992–2003 Documenting the scope and scale of FBO activities documented in studies

DeHaven, Hunter, Wilder, Walton, & Berry —Health Programs in Faith-Based Organizations: Are They Effective?

2004 53 28 1990–2000 Health programs delivered in churches or by health ministries

Ferguson, Wu, Spruijt-Metz, & Dyrness—Outcomes Evaluation in Faith-Based Social Services: Are We Evaluating Faith Accurately?

2007 29 29 1987–2004 Program effectiveness and faith-based organizations

Note. FBCO=Faith-Based and Community Organizations; FBO=Faith-Based Organization

APPENDIX C

Systematic Review Methods

Study Domains Not Included in Present Review

1. Studies of the relationship between religiosity and the status of individuals. These studies examine the relationship between religious behaviors (e.g., church attendance), beliefs, and attitudes, and the occurrence of a range of positive and negative behaviors. Research on what Johnson et al. (2002) termed “organic” religion, demonstrates that religious participation and belief are related to a wide range of positive social and health outcomes (Johnson, 2002; Johnson & Siegel, 2003; Powell, Shahabi, & Thoresen, 2003; Wilcox, 2002). There is also a

195

strong correlational link showing that youths with more religious involvement show better academic progress (Regnerus, 2002), less juvenile delinquency (Johnson, De Li, Larson, & McCullough, 2000), less criminal activity (Baier & Wright, 2001), and are more likely to engage in a range of healthy behaviors (Wallace, 2002). These studies provide important context, but they do not speak to the effects of particular program initiatives.

2. Studies of programs targeting church congregations. These so-called “faith-placed” efforts use the church setting as a venue for delivering an intervention, targeting members of a congregation for programs often delivered by other members. These studies often involve services delivered to the members of a congregation, not to a general target population in need. The challenge is that congregation members are often considered the active agent of service provision rather than the recipients of service. Congregations are excluded because they do not reflect the primary theoretical frame underlying the FBCI that involves offering services to a community population that is not simply the membership of a faith community. For example, studies have examined the effectiveness of offering church-based breast and cervical cancer screening interventions (Matthews, Berrios, Darnell, & Calhoun, 2006), automobile restraint use (Falcone, Brentley, Ricketts, Allen & Garcia, 2006), breast self-examination and mammography promotion (Erwin, Spatz, Stotts, & Hollenberg, 1999), fruit and vegetable consumption (Campbell et al., 1999), and smoking cessation (Schorling et al.,1997). Although it is clear that the vast majority of congregations do self-identify as offering one or more types of social service programming (e.g., Cnaan & Boddie, 2001), these programs are often quite distinct from services offered by FBCOs. Some scholars (e.g., Jeavons, 2004) have argued that, and for this and other conceptual reasons these studies are not reviewed.

3. Studies that are descriptive in nature or use only qualitative methods. Numerous studies have examined the approach and theory underlying FBCO interventions or have focused exclusively on assessing the implementation of these services. For example, Hodge and Pittman (2003) examined the characteristics of a sample of faith-based drug and alcohol treatment providers in Texas. This type of work is critically important for a number of reasons, including the essential clarification of program logic and program delivery models. This clarity will aid in the effective evaluation of these programs and serve to inform program development efforts across the board (Fischer & Stelter, 2006). However, such studies do not measure program benefit or document participant improvement. While it is true that FBCO social service programs contain many of the same primary programmatic

196

elements as secular programs, these faith-based elements should be clearly understood, quantified, and assessed as the field moves forward.

Effect Size Calculations

The formulation of the effect size metric is based on Cohen's (1969) original notion of standardized mean difference (d statistic) between two groups on an outcome of interest. In a treatment-effectiveness synthesis, the effect size is characteristically calculated as the raw difference in means for two groups divided by a measure of variation (e.g., the standard deviation) of the outcome distribution. Among these studies, group-level standard deviations were rarely reported, so the pooled variances were estimated using a procedure based on sample size (Raudenbush, 1994). The effect size can be calculated indirectly from reported significance test information for a given outcome. Other related transformations and manipulations are available for converting various reported significance and outcome information into the d statistic (Rosenthal, 1994).

Two calculations were necessary for the core meta-analytic procedure. First, a measure of impact (e.g., difference between treatment and control groups) was needed as the basic unit of the synthesis (d). Second, a sampling error variance estimate (v) was needed for each effect size estimate. The majority of the studies provided success rates for both groups, and these data were converted to an effect size using Cohen’s h statistic, an arcsine transformation (Lipsey & Wilson, 2001). Estimated effect sizes were computed for the remaining studies using whatever statistical data the authors presented in the original report. In the case of studies with multiple outcomes, the average effect size of the reported measures is presented in the Table D-1 (see Appendix D).

Limitations

The quantitative synthesis procedure (i.e., meta-analysis) is a useful tool but it does have a number of limitations. The external validity of research synthesis is important for the purpose of generalizing the review findings beyond the current set of studies. All reviews are threatened by bias due to the selectivity of the studies that are included. One particular concern is that meta-analysis is "dependent on the findings that researchers report" and bias will result if there are systematic differences among the results of research depending on whether they are published or not. This concern arises from the idea that many study results remain in "file-drawers" (Rosenthal, 1979) because their authors did not seek publication or were rejected, presumably because the results showed nonsignificant effects. As a result, the research retrieved by a meta-analysis may have a systematic bias. The present review handled the "file-drawer" threat by making an extensive search of the published and unpublished outlets for research on FBCOs.

197

198

Ultimately, only 3 of the 18 studies were retrieved from peer-reviewed sources. As such, the concern about a bias due to the file-drawer effect is substantially reduced.

An additional limitation is that quantitative reviews are dependent on the focus, quantity, and quality of the component studies available. As such, a review can only examine areas of research that have been the focus of attention in a sufficient number of individual studies. Also, the limitations of the individual studies in regard to data quality, reliability of measures, sample attrition, etc., are also a factor in the results of the review. As such, a review is ultimately only as good as the studies it comprises. The present review attempts to focus on the relatively higher quality studies available by restricting inclusion to those studies that use a comparative research design of some type (versus single-group designs).

APPENDIX D

Table D-1. Comparative Outcomes Studies of FBCO Services

Study Target

Population Design and Sample Size

Intervention Description

Outcomes Measures Findings

Subgroup Analyses

Mean Effect Size/95% CI

Bicknese (1999) – Teen Challenge

Substance users (TC graduates in PA, MO, and CA)

Tx (59) vs. C (118); matched controls from national sample

Tx – Bible-based classes plus employment C – short-term inpatient program

Substance abuse, employment, criminality

FBO group significantly more abstinent from drugs/alcohol (71.2% vs. 55.1%), more held full-time jobs (89.8% vs. 41.4%), fewer with arrest (7.0% vs. 17.0%)

By type of substance used

.581 (.558 – .604)

Campbell (2004) – Charitable Choice

Welfare clients in North Carolina

Tx (1,320) vs. C (10,862); Workforce Investment Act clients

Tx – faith-based employment services C – governmental employment services

Employment at program exit

Faith-based programs showed less employment success compared with governmental centers (40.9% vs. 79.2%)

By type of provider

−.807 (−.811 – −.809)

Deb & Jones (2003) – FaithWorks

Welfare clients, statewide in Indiana

Tx (2,930) vs. C (2,397); statistical controls

Tx – FaithWorks job training C – secular job training

Job placement, wages, hours, health insurance

Similar placement and wage rates; FBO clients work significantly fewer hours and fewer acquire jobs with health insurance

By gender, race, and high-school diploma status

.550 (.549 – .551)

Farley & Hackman (2006), Farley & McClanahan (2007) -Ready4Work

Former prisoners in 9 sites (6 faith-based, 3 secular)

Tx (2,374) vs. C (national sample – BJS)

Tx – employment readiness, job placement, mentoring C – usual services in communities

Recidivism at 6 months and 12 months

Participants had lower recidivism at 6 months (1.9% vs. 5.0%) and at 12 months (5.0% vs. 10.4%)

18- to 34-year-old African-American nonviolent felons

.190 (.189 – .191)

199

(continued)

Table D-1. Comparative Outcomes Studies of FBCO Services (continued)

Study Target

Population Design and Sample Size

Intervention Description

Outcomes Measures Findings

Subgroup Analyses

Mean Effect Size/95% CI

Florida Dept of Corrections (2000) – Kairos Horizons

Prisoners in Florida

Tx (67) vs. C (741); in Tomoka Prison

Tx – prisoners in faith-based dorm C – prisoners in other dorms

Disciplinary reports

Prisoners in faith-based program had fewer disciplinary reports (8.9% vs. 16.7%)

Completers vs noncompleters

.236 (.227 – .245)

Fox, Stein, Gonzalez, Farrenkopf, & Dellinger (1998)

Latino women in two communities in southern California

Tx (176) vs. C (126); samples in two communities 1990 and 1998

Tx – faith-based educational programming and mammography services C – usual services

Mammography use in previous year

Differences in mammography awareness (+19% vs. +12%); significant improvement in Tx community in receipt of mammography (27% vs. 24%)

Spanish speaking vs non-Spanish speaking

.174 (.161 – .187)

Johnson (2002) – Humaita

Prisoners in two prisons in Brazil

Tx (247) vs. C (148); matched sites

Tx – Prison staffed by church volunteers C – Prison operated by nonprofit

Recidivism over 3 years

FBO prisoners had significantly lower recidivism (16% vs. 36%)

High-risk vs low-risk

.464 (.454 – .474)

Johnson & Larson (2003) – InnerChange Freedom Initiative

Prisoners in Texas

Tx (177) vs. C (1,754); matched group in Texas

Tx -In-prison bible-based programming and 6-12 months of aftercare C – usual services

Rearrest and incarceration over 2 years

No significant difference on rearrest (36.2% vs. 35%) or incarceration (24.3% vs. 20.3%)

Completers vs noncompleters

−.061 (−.065 – −.057)

Johnson, Larson, & Pitts (1997); Johnson (2004) – Prison Fellowship

Prisoners, in four New York prisons

Tx (201) vs. C (201); matched group

Tx – Prison Fellowship C – no intervention

Recidivism at 2 years, 3 years, 8 years

No significant differences between two main groups at 2 years (44% vs. 43%), 3 years (53% vs. 57%), or 8 years (68% vs. 73%)

Level of program participation, level of risk

.057 (.047 – .067)

200

(continued)

Table D-1. Comparative Outcomes Studies of FBCO Services (continued)

Study Target

Population Design and Sample Size

Intervention Description

Outcomes Measures Findings

Subgroup Analyses

Mean Effect Size/95% CI

Jucovy (2003) – Amachi

Children of prisoners in Philadelphia

Tx (399) vs. universe of BBBS mentor matches

Tx – Volunteer mentors from churches C – Standard mentors

Mentor relationship lasting more than one year

Program had higher rate of matches lasting one year or more (62% vs. 46%)

n/a .322 (.317 – .327)

LaVigne, Brazzell, & Small (2007) – Faith & Character-based Institutions

Prisoners in Florida

Tx (289) vs. C (289); matched group from waitlist and general prison population

Tx – variety of faith-based programs delivered by nonprofits C – no special programming

Recidivism over 6 and 12 months

Significantly lower recidivism for males at 6 months (0% vs. 2.1%) but not at 12 months (1.8% vs. 2.4%); nonsignificant difference for females at 6 months (0% vs. 1%) and 12 months (1.9% vs. 6.5%)

Males vs females

.167 (.160 – .174)

Modesto (2006) – Welfare-to-Work

Welfare recipients in 3 North Carolina counties

Tx (102) vs. C (78); matched

Tx – faith-based program plus community college program C – secular program +community college

Earnings, poverty, employment

No significant differences between the groups on the measures available

County of residence, race

.383 (.361 – .405)

Monsma & Soper (2003, 2006); Monsma (2006) – Welfare-to-Work

Welfare recipients in Los Angeles

Tx (102) vs. C1 (141), C2 (113), C3 (80); clients served at 17 agencies

Tx – welfare-to-work program from FBO C – welfare-to-work program from government, for-profit, or secular agency

Employment (6/12 months), wages, TANF receipt

FBO programs more successful in helping employed clients retain a job (~90% vs. 61%–77%) but less successful in getting unemployed clients employed (~24% vs. 42%–60%)

By 5 provider types

−.015 (−.026 – −.004)

201

(continued)

Table D-1. Comparative Outcomes Studies of FBCO Services (continued)

Study Target

Population Design and Sample Size

Intervention Description

Outcomes Measures Findings

Subgroup Analyses

Mean Effect Size/95% CI

O’Connor, Su, Ryan, Parikh, & Alexander (1997) – Transition of Prisoners (TOP)

Former prisoners in Detroit

Tx (95) vs. C1 (88) waitlist, C2 (85) matched group

Tx – church-based mentor and support C – no intervention

Recidivism over 3 years

FBO program had lower rate of return to prison for escape (33% vs. 43%, 57%); no difference in recidivism due to parole violation or new crime (33% vs. 21%, 34%)

By program participation level, education level, prior felony

.254 (.239 – .269)

Ragan (2004a) –Nursing Homes

Nursing home clients, national sample

Tx (948) vs. C (15,342); nursing homes

Tx – religiously affiliated homes C – secular homes

Inspection deficiencies, complaint deficiencies

FBO homes had significantly lower rates of inspection deficiencies (4.40 vs. 5.93) and complaint deficiencies (1.46 vs. 3.44)

n/a .193 (.192 – .194)

Ragan (2004b) – Home Health Agencies

Home health clients, national sample

Tx (445) vs. C (6,723); home health agencies

Tx – religiously affiliated home health agencies C – secular agencies

11 measures of patient outcomes

8 of 11 measures showed that patients served by religiously affiliated agencies improved significantly more

n/a .031 (.030 – .032)

Stahler, Kirby, & Kerwin (2007) – Bridges program

Cocaine- dependent women in Philadelphia

Tx (8) vs. C (10); random assignment

Tx – residential program, faith-based mentoring and group activities C – residential program and secular group activities

Substance use at 3 and 6 months

No significant difference in abstinence at 3 months (88% vs. 66%), but at 6 months FBO program had significantly higher rates of abstinence (75% vs. 30%)

n/a .736 (.517 – .955)

202

(continued)

203

Table D-1. Comparative Outcomes Studies of FBCO Services (continued)

Study Target

Population Design and Sample Size

Intervention Description

Outcomes Measures Findings

Subgroup Analyses

Mean Effect Size/95% CI

Wilson, Picheca, & Prinzo (2005) – Circle of Support and Accountability (COSA)

Sex offenders in Ontario, Canada

Tx (60) vs. C (60); matched group

Tx – Participants assigned to 5 community volunteers who form a support group C – No intervention

Recidivism – sexual offense, violent offense

Program group had significantly fewer sexual offenses (5.0% vs. 16.7%) and violent offenses (15.0% vs. 35.0%) over 4.5 years

n/a .431 (.398 – .464)

Note: In the design and sample and intervention description columns, Tx refers to the treatment condition and C refers to the comparison condition. The Ragan (2004) report contains two distinct studies, which are listed separately here as (2004a) and (2004b).

APPENDIX E

Figure E-1. Effect Sizes for Studies Included in the Analysis

1 Bicknese (1999) 0.5812 Campbell (2004) -0.8103 Deb & Jones (2003) 0.5504 Farley & Hackman (2006) 0.1905 FL Dept of Corrections (2000) 0.2366 Fox et al. (1998) 0.1747 Johnson (2002) 0.4648 Johnson & Larson (2003) -0.0619 Johnson et al. (1997) 0.057

10 Jucovy (2003) 0.32211 LaVigne et al. (2007) 0.16712 Modesto (2006) 0.38313 Monsma & Soper (2003, 2006) -0.01514 O'Connor et al (1997) 0.25415 Ragan (2004a) 0.19316 Ragan (2004b) 0.03117 Stahler et al. (2007) 0.73618 Wilson et al. (2005) 0.43119 Weighted Mean (d) 0.205

0.431

0.736

0.031

0.193

0.254

-0.015

0.383

0.167

0.322

0.057

-0.061

0.464

0.174

0.236

0.190

0.550

-0.810

0.581

-1 -0.5 0 0.5 1Effect Size (d)

Note: The Ragan (2004) report contains two distinct studies, which are listed separately here as (2004a) and

(2004b).

REFERENCES

Baier, C. J., & Wright, B. R. E. (2001). If you love me, keep my commandments: A meta-analysis of the effect of religion on crime. Journal of Research in Crime and Delinquency, 38(1), 3–21.

Bartowski, J. P., Call, V. R. A., Heaton, T. B., & Forste, R. (2007). Religion, job readiness, and employment outcomes: The case of Latter-Day Saint Employment Resource Services. Research on Social Work Practice, 17(2), 188–198.

Berk, R. A. (1983). An introduction to sample selection bias in sociological data. American Sociological Review, 48, 386–398.

Bicknese, A. T. (1999). The teen challenge drug treatment program in comparative perspective. Unpublished doctoral dissertation, Evanston, IL: Northwestern University.

204

Branch, A.Y. (2002). Faith and action: Implementation of the national faith-based initiative for high risk youth. Philadelphia: Public/Private Ventures. Retrieved December 10, 2004, from http://www.ppv.org

Camp, S. D., Klein-Saffran, J., Kwon, O., Daggett, D. M., & Joseph, V. (2006). An exploration into participation in a faith-based prison program. Criminology & Public Policy, 5(3), 529–550.

Campbell, D. (2004). CFBO roles in serving the hard-to-employ: A government-led partnership to realize public value. Retrieved March 7, 2008 from http://groups.ucanr.org/CCP/files/34394.pdf

Campbell, D., Glunt, E., Bockman, S., Sirotnik, B. W., & Little, J. (2003, March). Evaluating the California community and faith-based initiative. Paper presented at the 2003 Spring Research Forum of the Independent Sector, Bethesda, MD.

Campbell, M. K., Denmak-Whnefried, W., Symons, M., Klasbeek, W. D., et al. (1999). Fruit and vegetable consumption and prevention of cancer: The Black Churches United for Better Health Project. American Journal of Public Health, 89(9), 1390–1396.

Chaves, M. (2002). Religious organizations: Data resources and research opportunities. American Behavioral Scientist, 45(10), 1523–1549.

Clerkin, R. M., & Gronbjerg, K. A. (2007). The capacities and challenges of faith-based human service organizations. Public Administration Review, 67(1), 115–126.

Cnaan, R. A., & Boddie, S. C. (2001). Philadelphia census of congregations and their involvement in social services delivery. Social Services Review, 75(4), 559–580.

Cnaan, R. A., & Boddie, S. C. (2006). Setting the context: Assessing the effectiveness of faith-based social services. Journal of Religion & Spirituality in Social Work, 25(3/4), 5–18.

Compassion Capital Fund National Resource Center. (2005). Measuring outcomes. Intermediary development series. Washington, DC: Department of Health and Human Services. Retrieved from http://www.acf.hhs.gov/programs/ccf/resources/gbk_pdf/om_gbk.pdf

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Cook, T. D., & Campbell, D. T. (1979). In Quasi-experimentation: Design and analysis issue for field settings. Boston: Houghton Mifflin.

Cordray, D. S., & Fischer, R. L. (1994). Synthesizing evaluation findings. In J. S. Wholey, H. P. Hatry, & K. E. Newcomer (Eds.), Handbook of practical program evaluation (pp. 198–231).San Francisco: Jossey-Bass.

205

Deb, P., & Jones, D. (2003). Does faith work? A preliminary comparison of labor market outcomes of job training programs. In S. S. Kennedy & W. Bielefeld (Eds.), Charitable Choice: First results from three states (pp. 57–64), Indianapolis, IN: Center for Urban Policy and the Environment, School of Public and Environmental Affairs, Indiana University-Purdue University Indianapolis.

DeHaven, M. J., Hunter, I. B., Wilder, L., Walton, J. W., & Berry, J. (2004). Health programs in faith-based organizations: Are they effective? American Journal of Public Health, 94, 1030–1036.

DiIulio, J. J. (2000, Fall). Godly people in the public square. The Public Interest, 141.

Ebaugh, H. R., Pipes, P. F., Chafetz, J. S., and Daniels, M. (2003). Where’s the religion? Distinguishing faith-based from secular social service agencies. Journal for the Scientific Study of Religion, 42(3), 411–426.

Erwin, D. O., Spatz, T. S., Stotts, R. C., & Hollenberg, J. A. (1999). Increasing mammography practice by African American women. Cancer Practice, 7(2), 78–85.

Fagan, P. F., Horn, C., Edwards, C. W., Woods, K. M., & Caprara, C. (2007, March). Outcome-based evaluation: Faith-based social service organizations and stewardship. Washington, DC: The Heritage Foundation.

Falcone, R. A., Brentley, A. L., Ricketts, C. D., Allen, S. E., & Garcia, V. F. (2006). Development, implementation and evaluation of a unique African-American faith-based approach to increase automobile restraint use. Journal of the National Medical Association, 98(8), 1335–1341.

Farley, C. & Hackman, S. (2006). Ready4Work in brief. Interim Outcomes are in: Recidivism at half the national average. In Brief, 4. Philadelphia: Public/Private Ventures.

Farley, C., & McClanahan, W. S. (2007). Ready4Work in brief. Update on outcomes; Reentry may be critical for states, cities. In Brief, 4. Philadelphia: Public/Private Ventures.

Ferguson, K. M., Wu, Q., Spruijt-Metz, D., & Dyrness, G. (2007). Outcomes evaluation in faith-based social services: Are we evaluating faith accurately? Research on Social Work Practice, 17(2), 264–276.

Fischer, R. L. (2001). The sea change in nonprofit human services: A critical assessment of outcomes measurement. Families in Society, 82, 561–568.

Fischer, R. L. (2004). The devil is in the details: Implementing outcome measurement in faith-based organizations. Nonprofit Management & Leadership, 15(1), 25–40.

Fischer, R. L., & Stelter, J. D. (2006). Testing faith: Improving evaluation research on faith-based human services. Journal of Religion & Spirituality in Social Work, 25(3/4), 105–122.

206

Fink, B., & Branch, A. Y. (2005). Promising practices for improving the capacity of faith and community-based organizations. Prepared for Administration of Children and Families, U.S. Department of Health and Human Services. May.

Florida Department of Corrections. (2000, September). Comparing Tomoka Correctional Institution’s faith-based dorm (Kairos Horizons) with non-participants. Tallahassee, FL: Bureau of Research and Data Analysis.

Fox, S. A., Stein, J.A., Gonzalez, R. E., Farrenkopf, M., & Dellinger, A. (1998). A trial to increase mammography utilization among Los Angeles Hispanic women. Journal of Health for the Poor and Underserved, 9, 309–321.

Hangley, B., & McClanahan, W. S. (2002). Mustering the armies of compassion in Philadelphia: An analysis of one year of literacy programming in faith-based institutions. Philadelphia: Public/Private Ventures. Retrieved December 10, 2004, from http://www.ppv.org

Hatry, H., Van Houten, T., Plantz, M., & Greenway, M. (1996). Measuring program outcomes: A practical approach. Alexandria, VA: United Way of America.

Hodge, D. R., & Pittman, J. (2003). Faith-based drug and alcohol treatment providers: An exploratory study of Texan providers. Journal of Social Service Research, 30(1), 19–40.

Independent Sector. (2003). The role of faith-based organizations in the social welfare system— A report on the 2003 spring research forum. Retrieved September 3, 2003, from http://www.independentsector.org/PDFs/srf03report.pdf

Jeavons, T. H. (2004). Religious and faith based organizations: Do we know one when we see one? Nonprofit and Voluntary Sector Quarterly, 33, 140–145.

Johnson, B. R. (2002). Assessing the impact of religious programs and prison on recidivism: An exploratory study. Texas Journal of Corrections, 28, 7–11.

Johnson, B. R. (2004). Religious programs and recidivism among former inmates in prison fellowship programs: A long-term follow-up study. Justice Quarterly, 21(2), 329–354.

Johnson, B. R., De Li, S., Larson, D. B., & McCullough, M. (2000). A systematic review of the religiosity and delinquency literature. Journal of Contemporary Criminal Justice, 16(1), 32–52.

Johnson, B. R., & Larson, D. B. (2003, June). The Innerchange Freedom Initiative: A preliminary evaluation for a faith-based prison program. Report from the Center for Research on Religion and Urban Civil Society, University of Pennsylvania. Philadelphia: Center for Research on Religion and Urban Civil Society.

Johnson, B.R. Larson, D.B. & Pitts, T.C. (1997). Religious programs, institutional adjustment, and recidivism among former inmates in prison fellowship programs. Justice Quarterly, 14(1), 145–166.

207

Johnson, B. R., & Siegel, M. B. (2003). The great escape—How religion alters the delinquent behavior of high-risk adolescents. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

Johnson, B. R., Tompkins, R. B., & Webb, D. (2002). Objective hope—Assessing the effectiveness of faith-based organizations: A review of the literature. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

Johnson, C. L. (2000). The Imani Project: Challenging missionary Baptist men to mentor adolescent African-American males. Unpublished doctoral dissertation, Drew University.

Jucovy, L. (2003). Amachi: Mentoring children of prisoners in Philadelphia. Philadelphia: Public/Private Ventures and Branch Associates. Retrieved December 10, 2004, from http://www.ppv.org/ppv/publications/publications.asp?search_id=3

Kennedy, S. S., & Bielefeld, W. (2003). Charitable Choice: First results from three states. Indianapolis, IN: Center for Urban Policy and the Environment.

LaVigne, N. G., Brazzell, D., & Small, K. (2007). Evaluation of Florida’s faith- and character-based institutions. Final report. Washington, DC: The Urban Institute.

Leake, R., Green, S., Marquez, C., Vanderburg, J., Guillaume, S., & Gardner, V.A. (2007). Evaluating the capacity of faith-based programs in Colorado. Research on Social Work Practice, 17(2), 216–228.

Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.

MacMaster, S. A., Jones, J. L., Rasch, R. F. R., Crawford, S. L., Thompson, S., & Sanders II, E. C. (2007). Evaluation of a faith-based culturally relevant program for African-American substance users at risk for HIV in the southern United States. Research on Social Work Practice, 17(2), 229–238.

Matthews, A. K., Berrios, N., Darnell, J. S., & Calhoun, E. (2006). A qualitative evaluation of a faith-based breast and cervical cancer screening intervention for African-American women. Health Education & Behavior, 33(5), 643–663.

Mears, D. P., Roman, C. G., Wolff, A., & Buck, J. (2006). Faith-based efforts to improve prisoner reentry: Assessing the logic and evidence. Journal of Criminal Justice, 34, 351–367.

Modesto, K. F. (2006). Not by faith alone: A preliminary outcomes evaluation of a faith-based welfare-to-work program. Doctoral dissertation submitted to the faculty of the University of North Carolina at Chapel Hill.

Monsma, S. V., & Mounts, C. M. (2002). Working faith: How religious organizations provide welfare-to-work services. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

208

Monsma, S. V., & Soper, J. C. (2003, September). What works: Comparing the effectiveness of welfare-to-work programs in Los Angeles. Philadelphia, PA: Center for Research on Religion and Urban Civil Society.

Monsma, S. V., & Soper, J. C. (2006). Faith hope & jobs: Welfare-to-work in Los Angeles. Washington, DC: Georgetown University Press.

O’Connor, T., Su, Y., Ryan, P., Parikh, C., & Alexander, E. (1997). Detroit transition of prisoners: Final evaluation report. Washington, DC: Prison Fellowship Ministries.

Powell, L. H., Shahabi, L., & Thoresen, C. E. (2003) Religion and spirituality: Linkages to physical health. American Psychologist, 58(1), 36–52.

Ragan, M. (2004). Faith-based vs. secular: Using administrative data to compare the performance of faith-affiliated and other social service providers. The Roundtable on Religion and Social Welfare Policy, Retrieved May 9, 2005, from http://www.religionandsocialpolicy.org/docs/research/Benchmarking_report_12-23-04.pdf

Raudenbush, S. W. (1994). Random effects models. In H. Cooper & J. V. Hedges (Eds.), The handbook of research synthesis (pp. 301–321). New York: Russell Sage Foundation.

Regenerus, M. D. (2002). Making the grade—The influence of religion upon the academic performance of youth in disadvantaged communities. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

Rock, J. (2002, October). Stepping out on faith: New York City’s Charitable Choice demonstration program. Albany, NY: SUNY, Rockefeller Institute of Government, The Roundtable on Religion and Social Welfare Policy.

Roman, C. G., Wolff, A., Correa, V., & Buck, J. (2007). Assessing intermediate outcomes of a faith-based residential prisoner reentry program. Research on Social Work Practice, 17(2), 199–215.

Rosenthal, R. (1979). The “file drawer problem” and tolerance for null results. Psychological Bulletin, 86, 638–641.

Rosenthal, R. (1994). Parametric measures of effect size. In H. Cooper & J. V. Hedges (Eds.), The handbook of research synthesis (pp. 231–244). New York: Russell Sage Foundation.

Schorling, J. B., Roach, J., Siegel, M., Baturka, N., et al. (1997). A trial of church based smoking cessation interventions for rural African-Americans. Preventive Medicine, 26, 92–101.

Scott, J. D., Montiel, L. M., Keyes-Williams, J., & Han, J. S. (2003, June). The scope and scale of faith-based social services: A review of the research literature focusing on the activities of faith-based organizations in the delivery of social services (2nd ed.). Albany, NY: SUNY, Rockefeller Institute of Government, The Roundtable on Religion and Social Welfare Policy.

209

Sherman, A. L. (2006). Strengthening the street saints: Intermediaries providing capacity-building assistance to faith-based organizations. In R.V. Anglin (Ed.), Building the organizations that build communities: Strengthening the capacity of faith- and community-based development organizations (pp. 73–88).Washington, DC: Office of Policy Development and Research, U.S. Department of Housing and Urban Development.

Sider, R. J., & Unruh, H. R. (2004). Typology of religious characteristics of social service and educational organizations and programs. Nonprofit and Voluntary Sector Quarterly, 33(1), 109–134.

Singer, P., & Friel, B. (2007, January 5). Leaps of faith. National Journal. Retrieved February 28, 2008, from http://nationaljournal.com/about/njweekly/stories/2007/0105nj1.htm

Smith, E. D. (1992). Hypertension management with church-based education: A pilot study. Journal of the National Black Nurses Association, 6, 19–28.

Smith, S. R., Bartkowski, J. P., & Grettenberger, S. (2005). Comparative views on the role and effect of faith in social services. Albany, NY: SUNY, Albany Rockefeller Institute of Government, The Roundtable on Religion and Social Welfare Policy.

Smith, S. R., & Sosin, M. R. (2001). The varieties of faith-related agencies. Public Administration Review, 61, 651–670.

Stahler, G. J., Kirby, K. C., & Kerwin, M. E. (2007). A faith-based intervention for cocaine-dependent black women. Journal of Psychoactive Drugs, 39(2), 183–190.

Twombly, E. C. (2002). Religious versus secular human service organizations: Implications for public policy. Social Science Quarterly, 83(4), 947–961.

U.S. Department of Health and Human Services. (2002, June). The Compassion Capital Fund and the Faith- and Community-based Initiative, HHS fact sheet. Retrieved September 4, 2003, from http://www.hhs.gov/news/press/2002pres/20020605.html

U.S. General Accounting Office. (2002). Charitable Choice: Federal guidance on statutory provisions could improve consistency of implementation. GAO-02-887. Washington, DC: author.

U.S. Government Accountability Office. (2006). Faith-Based and Community Initiative: Improvements in monitoring grantees and measuring performance could enhance accountablity. GAO-06-616. Washington, DC: author.

Wall Street Journal. (2003, June 20). Jesus saves: How President Bush found himself hugging a murderer in the White House. Editorial. Retrieved September 10, 2003, from http://www.opinionjournal.com/taste/?id=110003652.

210

Wallace, J. M. (2002). Is religion good for adolescent health? A national study of American high school seniors. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

White House. (2008, February). The Quiet Revolution—The President’s Faith-Based and Community Initiative: A seven-year progress report. Retrieved March 20, 2008, from http://www.whitehouse.gov/government/fbci/reports.html

Wilson, R. J., Picheca, J. E., & Prinzo, M. (2005, May). Circle of support and accountability: An evaluation of the pilot project in South-Central Ontario. Correctional Service of Canada.

Working Group on Human Needs and Faith-Based and Community Initiatives. (2002). Finding common ground: 29 recommendations of the working group on human needs and faith-based and community initiatives. Washington, DC: author.

Working Group on Human Needs and Faith-Based and Community Initiatives. (2003). Harnessing civic and faith-based power to fight poverty. Washington, DC: author.

211

FAITH-BASED INSTITUTIONS AS PROJECT IMPLEMENTERS: AN INNOVATIVE ECONOMIC EMPOWERMENT INTERVENTION

FOR CARE AND SUPPORT OF AIDS-ORPHANED AND VULNERABLE CHILDREN IN RURAL UGANDA

Fred M. Ssewamala Leyla Ismayilova

Among the world’s adults and children with HIV, 24.7 million (63%) live in sub-Saharan Africa (UNAIDS, 2006). Despite increased access to antiretroviral therapy in the region, 72% of all deaths are due to AIDS (UNAIDS, 2006). The AIDS pandemic has devastated the African continent and it has left millions of people in poverty and poor health. It has also left millions of children without their biological parents (UNICEF, 2003). Because of the large number of orphans that are overwhelming traditional African families, it is not uncommon to find instances where families are unwilling to take in additional orphaned children, even if they are their kin. This has led to an emerging trend of child-headed households, hitherto unknown in traditional African families and cultures.

The most recent statistics from Uganda indicate that 2.3 million children below age 18 are orphaned, having lost one or both parents (UNICEF, 2008). Of these documented orphans, about half (51%) are a direct result of the AIDS pandemic. Even with the falling HIV infection rates in the country, it is projected that the number of orphans will remain high, or even increase, as parents already infected with HIV die from the disease. Estimates indicate that one of every four households in Uganda is providing care to an orphaned child (Cheng, Hite, Jacob, & Smith, 2004).

Traditionally, the burden of raising orphans in Uganda—and most parts of sub-Saharan Africa—falls primarily on extended family members, including grandparents, uncles, and aunts (Ankrah, 1993; Foster, 2004; Matshalaga & Powell, 2002). However, the steady increase in the number of orphans, coupled with the increase in poverty, has contributed to a breakdown in the traditional extended family system, and there are no established public welfare programs that can help serve these children. Thus, a considerable number of orphans who would otherwise have been cared for through the extended family system have either dropped out of school in order to farm the land to take care of themselves and their siblings or, in desperation, have migrated to large urban areas in search of employment opportunities (Ssewamala & Curley, 2006). Because the majority of the orphans who migrate to the urban areas have no employable skills, they often end up on the streets, where they beg, engage in petty theft, begin drug and other substance abuse, and

213

prostitute themselves for money, exposing them to HIV infection and other sexually transmitted infections (Ssewamala, 2005). Uganda is one of many African countries where there is a noticeable increase in the number of child-headed households, indicating the lack of capacity and support structure within communities to care for vulnerable children (World Bank, 2000b).

Civil society institutions, including faith-based organizations, have initiated a number of programs and services for care and support of families and communities affected by HIV/AIDS. Over and above their traditional roles, faith-based organizations are becoming more actively involved in the provision of care services for orphaned and vulnerable children (Foster, 2002, 2005; Lee, Foster, Makufa, & Hinton, 2002) and implementation of HIV-related interventions (Kagimu, 2003; Liebowitz, 2002; Parry, 2003). This does not discount the fact that for generations faith-based and religious institutions have been involved in social service provision in one form or another in various parts of the world (Cnaan, Wineburg, & Boddie, 1999; Green, 2001; Lewis, 2003; Tangenberg, 2005). Thus, it is important to examine the interventions implemented by faith-based organizations, especially with an increase in funding coming to Africa under the President's Emergency Plan for AIDS Relief (PEPFAR) (Office of the Global AIDS Coordinator, 2004). For example, for a long time, faith-based organizations in sub-Saharan Africa were primarily associated with providing food, shelter, health care, and education; whereas recently, service delivery has changed and broadened. In Uganda, several faith-based organizations are actively involved in directly meeting the psychosocial needs of orphaned and vulnerable children, running social enterprise projects, and directly operating social infrastructure such as schools and hospitals (Kuchment, 2002). Among hospitals in Uganda, 40% are missionary hospitals with a faith-based connection (Kuchment, 2002), while over 30% of nongovernmental organizations (NGOs) in the country explicitly identify themselves as faith-based (Barr, Fafchamps, & Owens, 2005).

Prior studies have illustrated the relationship between asset-ownership and adolescents’ health and educational outcomes (Sherer et al., 2004; Zhan & Sherraden, 2003; Booysen & Van Der Berg, 2005). None of these studies, however, focused on care and support for adolescents through economic empowerment models—specifically microfinance-related models—implemented by a faith-based institution. Thus, this paper examines the efficacy and feasibility of an economic empowerment intervention implemented by a faith-based organization in rural Uganda. More specifically, the study aims to estimate the effects of the intervention on educational performance and aspirations, sexual risk-taking, and monetary savings behavior among orphaned and vulnerable children.1

214

The intervention presents an example of a nontraditional role for a faith-based organization in Uganda: collaborating with an academic institution (Columbia University) to rigorously research and test the efficacy and feasibility (including acceptability) of an economic empowerment intervention for care and support of orphaned and vulnerable children. In this particular case, the faith-based organization is not simply a recipient of aid from its collaborating institution (an American university), but an active participant in designing, implementing, monitoring, and evaluating an innovative intervention: the SEED/SUUBI2-Uganda project. This intervention presents an alternative approach for care and support of orphaned and vulnerable children by encouraging the children’s caregiving families to partner with faith-based organizations and local financial institutions to save money for educating orphaned and vulnerable children affected by HIV/AIDS.

AN ECONOMIC EMPOWERMENT INTERVENTION

Ensuring that the increasing number of children affected by HIV/AIDS can be cared for psychologically and economically continues to be a major challenge. Although Uganda is one of the countries highly affected by AIDS, the country still lacks a comprehensive operational national policy to support children orphaned or made vulnerable by the epidemic (for details see a joint report by UNAIDS, UNICEF, and USAID, 2004). Therefore, there is a compelling need for a program that incorporates the traditional care of marginalized populations—which tends to be provided by family and civil institutions, including faith-based institutions—with new interventions aimed at strengthening economic opportunities for families caring for orphaned and vulnerable children, developing the children’s future planning skills, enhancing their health functioning and educational opportunities, and reducing their risk-taking behaviors.

The SEED/SUUBI project goes considerably beyond existing orphan care, which primarily consists of institutionalization and reactive strategies in Uganda and other developing countries heavily affected by the AIDS pandemic. The intervention focuses on economic empowerment of families caring for orphaned and vulnerable children and addresses co-occurring health risks and poor educational achievements resulting from poverty and limited life options. Specifically, the intervention promotes children’s savings accounts, also known as Children’s Development Accounts (CDAs), specifically for postprimary education for orphaned and vulnerable children and microenterprise development (e.g., small income-generating businesses) for families caring for these children in Uganda.

The intervention is grounded in asset theory (Sherraden, 1990, 1991), which states that assets (e.g., savings, educational opportunities, economic opportunities in the form of income-

215

generating activities/microenterprises) have important economic, social, and psychological benefits for individuals and families. Asset-building—defined as efforts that enable people with limited financial and economic resources to acquire and accumulate long-term productive assets—is increasingly viewed as a critical factor for reducing poverty, positively impacting attitudes and behaviors, and improving psychosocial functioning (Ssewamala, Alicea, Bannon, & Ismayilova, 2008). Thus, an intervention aimed at orphaned and vulnerable children who live in poor households should include providing these children and their families or caregivers with economic empowerment opportunities through asset-building and asset-ownership. This can help children to envision the future with optimism, encourage planning for the future, and promote behavior change among those who might otherwise be vulnerable.

Implementation

SEED/SUUBI was implemented in two phases. The first phase was the SEED pilot study (2004–2006), which involved 100 orphaned and vulnerable children, with an average age of 13.6 years, who were selected from seven comparable primary schools located in the Masaka Diocese. The pilot project led to the second phase, the SUUBI Project, which was a larger study involving 286 children, with an average age of 14 years—this phase was funded by the National Institute of Mental Health (RFA # R21 MH076475-01). Hence, the current intervention name, SEED/SUUBI project. The ongoing collaboration with Masaka Diocese has been in existence since 2004.

The SEED/SUUBI project is based in St. Joseph’s Matale Parish, located in Rakai District, which is about 120 miles south of Kampala, the capital of Uganda. Rakai District is one of the districts hardest hit by HIV/AIDS and is also where the first AIDS case in Uganda was diagnosed 26 years ago (Serwadda et al., 1985). It is believed that two in five children in the district are likely to be orphaned. Currently St. Joseph’s Matale Parish—a faith-based institution (under Masaka Catholic Dioecese) located in Rakai district—is caring for or providing services to about 700 orphaned children, the majority of whom are orphaned because of AIDS. The parish provides counseling, support, and different kinds of aid to adults and children of HIV/AIDS-affected families. The children receive scholastic materials, school lunches, counseling services, and after-school programs (which include leisure and competitive sports and music activities). The parish also emphasizes the ABC model (Abstinence, Be Faithful, and Only use a Condom if you must) in its HIV prevention programs.

Across the two interventions that comprise the SEED/SUUBI project, a total of 386 children were involved in the study. The children were randomly selected from a total of 21 comparable

216

primary schools. The selected schools were all semi-urban public schools located in Rakai District, within Masaka Diocese. Each of the 21 primary schools was randomly assigned to the experimental or control condition such that all selected children from a particular school received the same intervention, primarily to address issues related to sample contamination. Each child in the control condition received the commonly used care for orphaned children (hereafter referred to as usual care), comprising the provision of recreation services, counseling, food aid (specifically school lunches), and scholastic materials (including text books) sponsored or administered by the Matale Parish.

In addition to the usual care mentioned above, the children in the experimental condition received an asset-based family intervention, which consisted of the following three components:

# workshops focused on asset-building and future planning;

# a monthly mentorship program with peers on life options, avoiding risk behavior, reinforcing learning, and building optimism; and

# a Child Development Account dedicated to paying for postprimary schooling and/or a family small business. This account provided the children with some financial resources with which they could begin to realistically plan for their future education or vocational training.

The argument behind providing the children with financial resources via a CDA—and the associated training—is rooted in the theory guiding the intervention. Asset theorists believe that if children do not see that they have the financial means with which to pursue long-term educational or vocational aspirations, then counseling or food aid may have little effect (Sherraden, 1986; Ssewamala et al., 2008). The one-to-one mentorship program is intended to help the orphaned children overcome a variety of challenges they face in daily life by fostering meaningful and lasting relationships with adult role models. The mentorship component is modeled on programs in the United States that have been found to improve child outcomes (Tierney, Grossman, & Resch, 1995).

The CDA is a matched savings account held in the child’s name in a recognized financial institution or bank. The child’s family members and friends are allowed, and indeed encouraged, to contribute to the CDA. The account is then matched with money from the program (i.e., intervention). The match cap—the maximum amount of family contribution to be matched by the intervention program—is the equivalent of US$10 a month per family. The match rate is 2:1, which means that if a child in the treatment condition had US$10 deposited each month in the CDA for 20 months (without withdrawals), at the end of the 20-month intervention period, this child would have a total of US$600 in the CDA, consisting of the US$200 family contribution

217

and US$400 in matched funds. This amount of money is enough to pay for an average of 3 years of a participant’s postprimary education in an ordinary public semi-urban school.

The children were restricted in the use of their matched savings to the following: (1) paying for education (postprimary school level or vocational training), and/or (2) investing in a family income-generating activity or small business (e.g., raising poultry, raising pigs for sale, buying a heifer/milk-producing cow). Each month, a savings account statement was generated for each child to see his or her accumulated savings and to serve as a morale booster for the participating families. Additionally, during the intervention period, each child (with his or her caregiver as a cosigner) had access to his or her own money in the account (excluding the matching funds), so that in case of an emergency the money could be withdrawn. The matching funds were kept in a separate account from a participant’s own savings and were not accessible to the participant. When a participant was ready to attend postprimary school, the check for the matching funds was written to the school the student chose to attend. The student then contributed his or her portion of the total cost for the academic term. The process was designed to eliminate the temptation of families to pressure the children to withdraw the money for their own use. It was also intended to avoid potential misuse of the matching funds by the children’s family members or caregivers.

METHODS

Measures

A 90-minute individual interview adapted from scales previously tested in the United States and South Africa (Auslander, Slonim-Nevo, Ozawa, Shepard, & Gehlert, 1992; Bhana et al., 2004; Levy, Lampman, Handle, Flay, & Weeks, 1993) was conducted with program participants at baseline and 12 months postintervention, assessing the effects of the SEED/SUUBI project on a wide range of children’s psychosocial functioning outcomes.

This paper specifically focuses on the SUUBI part of the study (N=286) and its effects on three outcomes: (1) educational planning and performance, (2) sexual risk-taking, and (3) savings (for the treatment group only because the control condition did not have verifiable savings).

Education Planning

In assessing a student’s future educational plans and aspirations, questions were adapted from the CHAMP Family Study Program in the United States and South Africa (Bhana et al., 2004; Levy et al., 1993; Paikoff, 1995). Children’s educational plans were measured using three questions. First, students were asked about their career interests: What do you want to be when you complete school? Further, children in both conditions were asked about their educational plans

218

after completing primary school (seventh grade). Finally, children who were planning to continue their education after secondary school (either through vocational training or an advanced level of education as a path to university) were asked to state how certain they were about achieving these educational plans. On this particular measure, the score ranged from 1 to 3, with a higher score indicating a higher degree of certainty.

Educational Performance

To measure educational performance, records were obtained from both the treatment and control schools on children’s school attendance and grades. Primary Leaving Examinations (PLEs) grades, which are standardized and nationally administered examinations taken at the end of primary grade seven, were used to measure children’s academic performance. PLEs were taken at least 7 months following the implementation of the project (i.e., the project was implemented in April and the PLE was taken in November of the same year). Because the study inclusion criteria for the schools were such that all schools to be included in the study had to have been at a comparable level of performance (as measured by the PLE grades) in the past 3 years prior to the intervention, coupled with the fact that children in the study (both in the control and the treatment condition) had relatively similar socioeconomic characteristics at baseline, any observable academic performance differences—as measured by a standardized national examination at 7 months postintervention—may to some extent be attributable to the intervention. Moreover, because the PLE grades used came directly from the schools and were confirmed with the students during the follow-up interviews, they are presumed to be accurate.

HIV and Sexual Risk-Taking

To assess intentions to engage in sexual risk behaviors, including those related to HIV/AIDS, questions were adapted from several sources (Auslander et al. 1992; Slonim-Nevo, Auslander, Munro, & Ozawa, 1994; Slonim-Nevo, Auslander, & Ozawa, 1995). Specifically, to measure children’s attitudes about abstinence, students were asked to assess two statements: (1) “it’s OK for someone to have premarital sex,” and (2) “my religion teaches me to wait until marriage.” Both items were measured on a 5-point scale (strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree). Further, students were asked about their attitudes toward abstinence as a way to avoid HIV (“Is not having sex the best way to avoid AIDS?”) and about their intention to abstain from sex to avoid HIV (“I do not plan to have sex because of HIV/AIDS”). Both items were measured on a scale from 1 to 4 (strongly disagree to strongly agree).

219

Savings

Only children in the treatment group received CDAs. The financial statements used for ascertaining the actual savings were obtained directly from the banks or financial institutions participating in the study. In this paper, savings is used as a measure of feasibility and to a small extent acceptability (although this is mainly captured through qualitative measures, the results of which are presented in Appendix A) of the intervention. The key research question related to the savings intervention is: Are these poor rural families able to use financial institutions specifically to save for the education of their orphaned and vulnerable children?

DATA ANALYSIS

Although the study used a mixed-method approach—combining both qualitative and quantitative measures—to evaluate the intervention, the principle focus of this paper is the quantitative findings. For a summary of emerging themes from the qualitative findings, see Appendix A.

Quantitative Data Analysis

Repeated Measures Analysis of Variance/ANOVA (or General Linear Model) was used to estimate the effect of the intervention comparing the treatment and control groups at baseline (Wave 1) and at 12 months postintervention (Wave 2). Wave 3 of data collection is ongoing and as such is not included in the present analysis. The analysis controlled for a child’s gender, age, and orphan status (single or double orphan, having lost one or both parents). The statistical analysis was performed using statistical package SPSS 15.0.

Who are SUUBI Project Beneficiaries? Sociodemographic Characteristics

As mentioned earlier, this paper specifically focuses on the SUUBI part of the study. Table 1 presents basic sociodemographic characteristics of the children involved in the projects. Out of 286 children involved in the SUUBI project, 57% were female. Participating families had, on average, six people in the household (ranging from two to 18 people). Besides the child who participated in the studies, families had, on average, three other children in the household ranging from 0 to 17 years of age. Almost 40% of participating children are double orphans. Primarily grandparents, aunts, and uncles were among the key caregivers. The majority of children in SUUBI self-identified as Catholic Christian (78.8%), 12% self-identified as Protestant Christian, 7.3% self-identified as Muslim, and 2.4% self-identified as Born-Again Christians. Overall, the community where the SUUBI project is being implemented—like many rural communities in Uganda—is fairly religious. About 81% of children willingly go to church or mosque almost every week and only a very small percentage (1%) never attend any religious services. It is important to note that participation in a specific religious activity or faith is not a

220

requirement for participation in the intervention. Children and families were recruited irrespective of their religious beliefs.

Table 1. Sociodemographic Characteristics of SUUBI Participants (N=286)

Key Characteristic Mean (SD) or %

(n=286)

Mean/SD or % Control Group

(n=148)

Mean (SD) or % Treatment Group

(n=138) Female child 57% 53% 60% Mean child age, years 13.7 (SD=1.3) 13.6 (1.47) 13.8 (1.35) Average # of people in household 6.7 (SD=3) 6.2 (SD=2.7) 7.1 (SD=3.2) Average # of children in household 3.5 (SD=2.5) 3.1 (SD=2.1) 3.9 (SD=2.8) Report father not living 81% 84% 78% Report mother not living 58% 58% 59% Report both parents not living 39% 41% 37% Female caregiver

Biological mother 40% 37% 42% Stepmother 3% 3% 3% Grandmother 36% 40% 32% Aunt 15% 14% 17% Sister 2% 3% 2% No female present 4% 4% 4%

Male caregiver Father 29% 24% 33% Grandfather 14% 13% 15% Uncle 11% 9% 15% Brother 4% 3% 4% No male present 41% 51% 33%

Religion Catholic 79% 79% 78% Protestant 12% 11% 12% Muslim 7% 9% 6% Born Again/Saved 2% 1% 4%

Religiosity/Frequency of attending church/mosque

Almost every week 81% 79% 83% Less than once a week, but more than just on holidays

16% 20% 13%

Only holidays 2% 1% 2% Almost never 1% 0% 2%

221

FINDINGS

Education-Related Outcomes

Educational Plans and Aspirations

The analysis indicates a significant effect of the SUUBI intervention on the educational plans of children. At baseline (also referred to as pretest), 78% of children in the control group reported that they planned to continue with their education and attend senior secondary schooling (equivalent to high school in the U.S. educational system). The percentage was nearly the same for children in the treatment group at baseline (77%). At 12-month follow-up, the number of children in the control group planning to go on to senior secondary schooling had dropped by 13 percentage points (from 78% to 65%). In contrast, the number of children in the treatment group planning to go on to secondary schooling had increased by eleven percentage points (from 77% to 88%). The results are statistically significant and the findings do not differ by gender. Similar results were reported in the pilot SEED project (Ssewamala et al., 2008; see also Appendix 2).

Career Preference/Options

The results from the SUUBI study indicate that students’ professional interests remained fairly unchanged in the control group. At baseline, about 59% of children in the control group indicated a preference for getting an academic degree and becoming an accountant, teacher, doctor, or lawyer. The percentage was considerably lower for the treatment group (38%). At 12-month follow-up, the percentage of children in the control reporting interest in academic degree-awarding professions/careers had dropped slightly by four percentage points to 55%, whereas the percentage for children in the treatment group had increased to 67%. There are several possible explanations for these changes. One reason could be that there is a sense of hope or belief among the treatment group children that they can continue their education and that they have the means (through the CDAs) to achieve their educational goals. In contrast, children in the control group probably could not envision how it would be possible to accomplish obtaining a career or profession requiring a degree without having the resources.

Certainty Level

For the treatment group, the results indicate a statistically significant change in the level of certainty children had in pursuing their educational goals. Specifically, a 21 percentage point increase (58% at baseline vs. 79% at 12-month follow-up) was registered in the treatment group children regarding their certainty of achieving their educational plans. In contrast, the percentage of children in the control group who were certain about achieving their educational plans dropped by eleven percentage points (62% at baseline vs. 51% at 12-month follow-up).

222

In regard to gender, the analysis indicated a statistically significant effect on girls in the treatment group compared with their male counterparts in the same group. Certainty score for girls in the treatment group rose by 28 percentage points (63% at baseline vs. 91% at posttest). The certainty change for boys in the treatment group increased by five percentage points (37% at baseline vs. 42% at posttest). These findings raise a key question: Why is the certainty change level for boys so much lower than that of their female counterparts? This might be at least partly explained by the fact that Uganda is a patriarchal society and, in such a setting, when children are orphaned, boys typically take on the responsibility of caring for the family and it is possible that most of the orphaned boys in the study were thinking more about taking care of their siblings and much less about going to school. However, given the current dataset, it is not possible to determine a precise explanation. A more qualitative measure may be necessary to address this question/observation.

Educational Performance

The PLE score can range from 4 to 36, with a lower score representing better academic achievement. Children in the treatment group demonstrated a better score on the national exam (26.82 average score) compared with their counterparts in the control condition (28.97 average score). The differences between the two groups are statistically significant.

In regard to gender, the results on educational performance indicate that girls in the treatment group scored, on average, 26.55 points on the PLE; whereas girls in the control group scored, on average, 29.58 points. The differences were statistically significant. There were, however, no statistical differences between the academic performance of boys in the treatment group (27.33 points) compared with the academic performance of boys in the control group (28.26 points).

School Attendance

The analysis indicated that at 12-month follow-up, compared with pretest, school attendance improved significantly in the treatment group and did not show statistically significant changes in the control group. In the treatment group, the number of days children attended school increased, on average, by 3 days (from 57.92 in the first term to 60.46 days in the second term). School attendance of children in the control group decreased slightly (from 62.7 in the first term to 61.85 in the second term). There were no significant differences by gender on school attendance following the intervention.

Attitudes about HIV and Sexual Risk-Taking

It is important to note that because the study uses an experimental research design with a treatment and control group, and because the two groups were drawn from very similar

223

socioeconomic and geographical circumstances at baseline (preintervention), the observable differences at the follow-up period (postintervention) may, to some extent, be attributable to the intervention.

Sexual Risk-Taking Behavior

At pretest only, three respondents (1%) reported having ever had sexual intercourse. All three children had had their first intercourse before age 15, and none of them reported using protection during the first intercourse. Although it is not possible to determine definitively, there may be several reasons as to why children reported low rates of sexual activity. One reason may be because of the age of respondents enrolled in the project. These are still relatively young children. Second, although devastated by disease, these are children from relatively stable rural communities. There could be some level of conservatism compared with children in an urban setting. Third, children might be underreporting their own sexual experiences, especially given the fact that 23% of the children in the study reported knowing one or more friends who had had sexual intercourse, and 18% reported feeling pressure to have sex. At 12-month follow-up, seven children reported having had intercourse in the past year (an increase of four children).

Attitudes are an important factor that could be used to predict future sexual risk-taking behavior (Raj, 1996; White, Terry, & Hogg, 1994). Because the numbers representing “actual” sexual risk-taking behaviors are small, the focus instead is on estimating the effects of the intervention on adolescents’ attitudes toward abstinence.

Attitudes Toward and Acceptance of Abstinence from Sex Until Marriage

During the follow-up period (posttest), there was an observable difference between students in the treatment group vis-à-vis students in the control group on “attitude and acceptance of abstinence from sex until marriage.” Adolescents in the control group had a mean score of 4.56 points (on the scale used to measure this construct) at pretest and 4.06 points at posttest—a reduction of 0.5 points in the mean score. In contrast, adolescents in the treatment group had a mean score of 3.73 points (on the same scale) at pretest and 4.35 points at posttest—an increase of 0.62 points. Because a high score represents a positive improvement in the attitudes toward abstinence until marriage, the results indicate that students in the treatment group demonstrated a significant improvement on this particular measure. The effect of intervention was statistically significant and was stronger for boys.

In regard to the second measure of abstinence from sex until marriage (specified as “my religion teaches me to abstain from sex”), there was no significant difference in the response to this

224

measure between the treatment group and control group, either at baseline or at postintervention follow-up.

Turning to the attitudes toward “abstinence as the surest way to avoid HIV,” the results indicate that at follow-up, the treatment group registered a more positive attitude toward abstinence as the best way to avoid HIV compared with their counterparts in the control group. The change is statistically significant. The treatment group’s score on this measure increased from 3.65 at baseline to 3.79 at 12-month follow-up, an increase of 0.14 points on this particular measure. In contrast, the control group reported slightly decreased results on the same measure: 3.87 at baseline as compared with 3.65 at 12-month follow-up, a reduction of 0.22 points on this measure.

In regard to gender, the results indicate that the intervention had a more significant effect on changing attitudes to abstinence among boys than girls. The results on this measure indicate that boys in the treatment group reported a more significant change in positive attitude toward abstinence as the best method of protection against HIV.

In behavioral science, there is a difference between attitudes and intentions. Usually intentions are regarded as a proxy for behavior, when behaviors are not easily measurable (Ajzen & Fishbein, 2005). Thus, this study aimed to not only establish the children’s attitudes to abstinence, but it also attempted to establish their actual intentions to “abstain from sex until marriage” to avoid HIV. On this measure—actual intentions to abstain from sex until marriage— students in the control group had a mean score of 3.50 points at pretest. At 12-month follow-up, the score had decreased by 0.30 points to 3.20 points. The change is not statistically significant. In contrast, the treatment students’ intentions to abstain from sex increased by 0.46 points (from 3.01 to 3.47). The results are statistically significant.

In regard to gender, the results point to statistically significant gender differences, suggesting that the intervention primarily had an effect on the intentions of boys and not girls. These results are similar to the findings in the SEED study (Ssewamala et al., 2008; Ssewamala, Bannon, Ismayilova, & Alicea, 2008).

Savings Outcomes

As mentioned earlier, savings as an outcome measure was only applied to the 138 children in the treatment condition because these are the children who had verifiable savings. Thus, the comparison group was not included in the analysis for this outcome measure, as children in the control condition did not have CDAs. The savings data indicated that participants in the

225

treatment condition (with CDAs) could and did save. Children in the treatment group in the SUUBI project saved, on average, US$6.33 a month or US$76 a year. There were no statistically significant differences in saving by gender and there were no statistically significant differences in savings by type of orphanhood. After matching individual savings—by a ratio of 2:1—the participants accumulated, on average, US$228 per year.

Data from the SEED pilot project (N=97) showed slightly better savings outcomes. In the first 6 months following the intervention, the participants saved an average of US$50.52 per participant or US$8.42 in average monthly deposits (Ssewamala et al., 2008).

Overall, although the savings amounts in SUUBI (a monthly average of US$6.33, or an accumulated yearly savings of US$228, including the match) may seem very modest by the standard of Western countries—these are huge sums in a poor country like Uganda, where annual per capita income is less than US$300 (World Bank, 2000a). Moreover, that amount of money is almost enough to pay for a student’s postprimary education for at least one year in an ordinary semi-urban public secondary school, which costs an average of US$200 per year.

IMPLICATIONS FOR RESEARCH, POLICY, AND PRACTICE

The SEED/SUUBI-Uganda intervention serves as an example of a well-planned collaboration between a U.S. academic institution and a religious/faith-based institution in rural Uganda. The results presented here suggest a potential for collaboration between a faith-based organization and a research institution in addressing care and support of a vulnerable group: orphaned and vulnerable children. Specifically, the results suggest that poor families caring for HIV/AIDS orphans in Uganda, if facilitated by trusted organizations, can save for the educational needs of the orphaned children they are caring for.

The results also suggest that savings—used as a measure of economic empowerment—might have an effect on the future of children by improving their expectations about future careers and increasing their motivation to make more careful/responsible choices regarding sexual risk-taking. If the future looks brighter, children may be more inclined to preserve it.

Finally, the qualitative work presented in Appendix A seems to suggests that collaboration with community-based organizations, specifically faith-based institutions that already have grounding and trust within the community, may play a role in the successful implementation of the projects aimed at care and support of orphaned and vulnerable children. It is not clear whether the results

226

would have been different if the project had been implemented through a governmental or nongovernmental secular organization; that is a question for future research.

Overall, experimenting with collaborations testing economic empowerment as an intervention for care and support of orphaned and vulnerable children in a poor developing country like Uganda may be worth considering and indeed funding.

ACKNOWLEDGMENTS

We are grateful to the SUUBI-Uganda Research Staff and volunteers for monitoring the study implementation process, especially Ms. Proscovia Nabunya, Reverend Fr. Kato Bakulu, and Ms. Stacey Alicea. We thank Professors Mary McKay, Jane Waldfogel, Michael Sherraden, and Nabila El-Bassel for helpful comments on the study intervention design, implementation, and data collection. Our thanks also go to all the children and their families or caregivers who agreed to participate in the SUUBI/SEED study. We appreciate the anonymous reviewers for their thoughtful comments. Financial support for the SUUBI-Uganda study came from the National Institute of Mental Health (RFA # R21 MH076475-01). Support for the pilot SEED study came primarily from the Center for Social Development at Washington University in St. Louis, the Social Intervention Group at Columbia University, and the Friedman Family Foundation.

NOTES

1. Because of space limitations, we have presented qualitative findings in the appendix of this paper, describing the role of a religious institution (in this paper also referred to as a faith-based organization) as a key project implementer, highlighting the strengths of the partnership vis-à-vis other available alternatives. The data are in the participants’ and local community members’ own words.

2. SEED (Savings for Entrepreneurship and Education) was a pilot project implemented in 2004–2006. SUUBI (which means “Hope” in the local language, Luganda) is an extension of the pilot project (2005–present).

APPENDIX A

Qualitative Findings: Process Evaluation

Qualitative responses were used to primarily assess acceptability of the intervention within the broader community. Specifically, in-depth interviews were conducted with four community leaders (two religious leaders and two political/government leaders); seven children and their caregivers from the treatment group and four children and their caregivers from the control group). For the qualitative analysis, children from the treatment group were purposively selected to represent different levels of saving: three children whose savings were above the average,

227

three children whose savings were at the average level, and one child whose savings were below average. Common themes in regard to the acceptability of the intervention and outcomes of interest were identified and quotes are presented. Overall, the qualitative results indicate the importance of collaboration to the success of the SUUBI/SEED initiative. The key components of the partnership are presented below along with the emerging themes.

Social Networks

Religious institutions and faith-based organizations often have established and long-standing contacts and networks within communities (Foster, 2004; Kagimu, 2003). The SUUBI project benefited substantially from the social networks and trust that communities have with the Masaka Diocese as a faith-based organization. The leaders of the diocese are familiar with community residents, their needs, and systems of services, such as schools and local (including district) counselors. For example, Masaka Diocese reverend fathers and pastors actively participated in identifying the schools to be included in the study, the potential participants/families from each of the schools, and the overall recruitment of the actual participants for the study. Pastors distributed the recruitment flyers during church services and during their visits to the community, informing willing and able caregivers for orphaned and vulnerable children within the community about the study.

In a study interview, the parish priest from St. Joseph’s Catholic Parish emphasized partnership and active involvement of the key stakeholders as important components of the project:

I think for the moment it [the project] is going on well… because we have got involvement of various partners, the main researchers, the coordinators of the project, the communities themselves and the kids are involved because there are general meetings which occur and we participate. And I’ve seen an interest especially on the part of the parents because now they are more involved. We meet them often…people coming to parish because of the project.

The priest went on to say:

The church we are fully involved in the community…we meet the kids in various spheres, we visit their schools, we visit their villages…we can usually mobilize the people…and that is a very big advantage we have over any other organization or maybe the government because we go deep into the villages…we live with them, we know their problems and we are more involved than maybe any other organization.

Trust and Cultural Integrity

Stemming from their long-term presence in the community, religious institutions have become integrated into the everyday lives of ordinary citizens. Community members in the study area

228

reported a feeling of “trust” they have in the religious institutions operating within their community. This is the same feeling SUUBI participants have toward the leadership of Masaka Diocese that introduced the project to the area. As Tyler (1997) rightly observed, if people trust, it means that they consider the person, or organization, they trust to be concerned about their needs and to “have their best interest at heart, to care about their views, to consider those views when decision-making, and to try to be fair” (p.2). In the eyes of the public, religious or faith-based organizations are always trying to look out for the best interest of the disadvantaged groups in communities. Thus, residents of the Masaka Diocese welcomed the intervention that was being offered through SUUBI primarily because they trusted Masaka Diocese, which introduced the intervention to the project area. The situation may have been different if the intervention had been implemented through different means, with no well-established presence or genuine trust from the community members.

As one of the project staff members explained:

We chose to work with the parish because people have trust in these priests at the parish. Like if they say something to the congregation, they’ll take it genuinely. They have trust in these people and also they reach out to these communities. Some of them know where some people live and they can always talk to them whenever they want. So that’s why we use the parish.

An Established Organizational Structure

The existing organizational structure within religious organizations allowed for a well-coordinated implementation of project activities involving a significant number of people/personnel across different locations. Masaka Diocese, through its parish priests, played an instrumental role in managing the project and coordinating all 21 project sites/schools located in over 14 different villages. For example, in addition to the recruitment process, the diocese coordinated all the research booster sessions with the children and their caregivers, and constantly reminded participants to use their savings accounts so they could take advantage of the matching funds. Further, the CDAs were monitored by the parish priest with the assistance of the study’s principal investigator at Columbia University and the project coordinator located in Uganda. The three had regular coordination meetings either in person or over the telephone. At the end of the savings period, the program personnel, together with representatives of Masaka Diocese, noted the savings in each participant’s account, subtracted the amount of money so far matched, and then ascertained the balance to be matched, assuring transparency and accountability in the community. Because the Masaka Diocese parish priests, the project coordinator, and the project assistants all were located in the community where the participating

229

families lived, they could monitor the process while consulting regularly with the principal investigator.

Mutual Benefits

Finally, the partnership was successful because the community felt that the project was able to offer something substantial to their community and they would not be exploited only for research purposes. This factor is especially important in the wake of so-called “research fatigue,” where communities are reluctant to be the subjects of research studies without substantial benefits to the community (Kasusse, 2005; Musoke, 2000; Thiessen et al., 2007).

As the Rakai District chairperson reported:

Quite a number of NGOs [nongovernmental organizations] have come since the early ‘90s … I thought this was an ordinary NGO. In fact, I was reluctant to come because I was busy. But I did come…and I found that SUUBI was slightly different. I liked their approach. They are giving our children support, but at the same time they are tying this support to the child himself participating himself or herself by saying that “if a child raises 2000 shillings, SUUBI will double that in contribution.” I thought this was an innovation that other NGOs could actually emulate…Very many people in Rakai when they look at NGOs they say “oh, they have brought us sugar, they have brought us milk, they will give us clothing” and now SUUBI is saying “no, we are going to help you but you also help yourself” and that is the way to go.

APPENDIX B

Table B-1. Summary Project Description

SEED/SUUBI-Uganda Project Description

Project Background The SEED/SUUBI-Uganda has been implemented since 2004 with the overall aim of testing the feasibility of an asset-based initiative with AIDS-orphaned children in Uganda. The project is a collaboration between Columbia University, New York (USA) and a faith-based organization called Masaka Catholic Diocese in Uganda.

The SEED/SUUBI-Uganda projects (a) test the feasibility and safety of conducting an asset-development intervention in developing country like Uganda, and (b) examine the effects of the intervention on children’s educational plans, health, and risk-taking behaviors.

Experiment Details Sample: 386 orphaned and vulnerable children aged 12 to 15 having lost one or both parents and enrolled in primary school (100 children participated in the pilot SEED project and 286 participated in SUUBI project).

Design: Experimental design with randomization administered at school level

(continued)

230

Table B-1. Summary Project Description (continued)

SEED/SUUBI-Uganda Project Description

Experiment Details (continued)

Project Sites: 21 semi-urban schools across 14 different villages around Rakai District in the South of Uganda.

Partner: Project is implemented through partnership with Masaka Diocese and St. Joseph’s Matale Parish.

Asset Intervention: In addition to the traditional care services provided to orphan children, an experimental arm, children received a family economic intervention, which included a Child Development Account (CDA) and six 2-hour classes on career planning, short-term and long-term career goals, microfinance, and financial well-being.

Accounts: Held in the adolescent’s name in a designated private financial institution— Centenary Rural Development Bank (a regulated MFI)—the CDA is funded by contributions from the adolescent’s family members or friends, with 2:1 matching funds from the intervention. Matching funds are provided by the project funders.

Uses: Account holders may use their personal savings in their CDA as they choose. However, they can only use the matched portion of the CDA to pay their educational expenses or to invest in such income-generating activities as raising livestock or starting a small business. Also, when they withdraw any savings for any reason other than these designated uses, they lose that relative portion of the match.

Impacts Health Outcomes: Children in the CDA experiment group had improved HIV prevention attitudes, whereas children in the control group showed decreased scores relative to baseline measurement on this variable. Boys demonstrated a more significant change compared with girls.

Savings: The CDA group could and did save. On average, participants saved US$75 per participant per year or US$6.33 per family in average monthly deposit. With a match rate of 2:1, the average participant accumulated an average of US$19 per month or US$228 per year. This is a huge amount in Uganda, where annual per capita income is less than US$300 and postprimary education costs about USD$200 per year.

Future Orientation: Data indicate that in addition to having money saved for their postprimary education, students in the CDAs group had a greater ability to identify specific future goals and aspirations, such as continuing their education.

Summary Preliminary results of both studies suggest that in a time when the HIV/AIDS epidemic in sub-Saharan Africa is destroying family and community networks, collaborations between various academic institutions like Columbia University and faith-based institutions in Rakai, Uganda, could result in successful partnerships to provide alternative care for orphan and vulnerable children while maintaining the cohesiveness of families and entire communities. The findings demonstrate promising outcomes in improving children’s educational aspirations, health outcomes, and social capital using an economic strengthening intervention.

231

REFERENCES

Ankrah E. M. (1993). The impact of HIV/AIDS on the family and other significant relationships: the African clan revisited. AIDS Care, 5(1), 5–22.

Auslander, W. F., Slonim-Nevo, V., Ozawa, M., Shepard, S. and Gehlert, S. (1992). Comparison of skills training, discussion, and control groups to prevent AIDS among abused and delinquent youth. VIII International Conference on AIDS/III STD World Congress, Amsterdam, The Netherlands, July 19–24.

Ajzen, I., & Fishbein, M. (2005). The influence of attitudes on behavior. In D. Albarracin, B. T. Johnson & M. P. Zanna (Eds.), The handbook of attitudes. Mahwah, N.J.: Lawrence Erlbaum Associates Publishers.

Barr, A., Fafchamps, M., & Owens, T. (2005). The governance of non-governmental organizations in Uganda. World Development, 33(4), 657–679.

Bhana, A., Petersen, I., Mason, A., Mahintsho, Z., Bell, & C., McKay, M. (2004). Children and youth at risk: Adaptation and pilot study of the CHAMP (Amaqhawe) programme in South Africa. African Journal of AIDS Research, 3(1), 33–41.

Booysen, F., & Van Der Berg, S. (2005). The role of social grants in mitigating the socio-economic impact of HIV/AIDS in two free state communities. South African Journal of Economics, 73, 545–63.

Cheng, S.Y., Hite, S. J., Jacob, W. J. & Smith, T. D. (2004). Helping Uganda’s street children: An analysis of the Model for Orphan Resettlement and Education (MORE). Journal of Children & Poverty, 10, 3–22.

Cnaan, R. A., Wineburg, R. J., & Boddie, S. C. (1999). The newer deal: Social work and religion in partnership. New York: Columbia University Press.

Foster, G. (2002). Understanding community responses to the situation of children affected by AIDS: Lessons for external agencies. Geneva: United Nations Research Institute in Social Development.

Foster, G. (2004). Study of the response by faith-based organizations to orphans and vulnerable children. UNICEF and World Conference of Religions for Peace.

Foster, G. (2005). Religion and responses to orphans in Africa. In Foster, G., Levine C., & Williamson, J. (Eds.), A generation at risk: The global impact of HIV/AIDS on orphans and vulnerable children. Cambridge: Cambridge University Press.

Green, E. C. (2001). The impact of religious organizations in promoting HIV/AIDS prevention. In Challenges for the church: AIDS, malaria & TB. Arlington, VA: Christian Connections for International Health.

232

Kagimu, M. (2003). Faith-based community mobilisation and education for antiretroviral therapy in Uganda. Sexual Health Exchange, 3, 12–14.

Kasusse, M. (2005). Bridging the digital divide in Sub-Saharan Africa: The rural challenge in Uganda. The International Information & Library Review, 37(3), 147–158.

Kuchmen, A. (2002, February 18). When it comes to AIDS, Africa's Christian church leaders have been whistling past the graveyard: Time to get religion. Newsweek. (International ed.).

Lee, T., Foster, G., Makufa, C., & Hinton, S. (2002). Families, orphans and children under stress in Zimbabwe. Evaluation and Program Planning, 25(4), 459–470.

Levy, S.R., Lampman, C., Handle, A., Flay, B.R., & Weeks, K. (1993). Young adolescent attitudes towards sex and substance abuse: Implications for AIDS prevention. AIDS Education and Prevention, 5, 340–351.

Lewis, B. M. (2003). Issues and dilemmas in faith-based social service delivery: The case of the Salvation Army of Greater Philadelphia. Administration in Social Work, 27(3), 87–106.

Liebowitz, J. (2002). The impact of faith-based organizations on HIV/AIDS prevention and mitigation in Africa. Health Economics and HIV/AIDS Research Division [HEARD], Durban, South Africa, University of Natal.

Matshalaga, N. R., & Powell, G. (2002). Mass orphanhood in the era of HIV/AIDS. British Medical Journal, 324(7331), 185–186.

Musoke, M. G. N. (2000). Information and its value to health workers in rural Uganda: A qualitative perspective. Health Libraries Review, 17(4), 194–202.

Office of the United States Global AIDS Coordinator (OGAC). (2004). The President's EmergencyPlan for AIDS Relief: U.S. five-year global HIV/AIDS strategy. Washington, DC: Office of the United States Global AIDS Coordinator. Retrieved from http://purl.access.gpo.gov/GPO/LPS51485

Paikoff, R.L. (1995). Early heterosexual debut: Situations of sexual possibility during the transition to adolescence. American Journal of Orthopsychiatry, 65(3), 389–401

Parry, S. (2004). Responses of the faith-based organizations to HIV/AIDS in Sub-Saharan Africa. Unpublished. Retrieved February 20, 2008, from http://www.synergyaids.com/documents/FBO_Responses_SSAfrica.pdf

Raj, A. (1996). Identification of social cognitive variables as predictors of safer sex behavior and intent in heterosexual college students. Journal of Sex and Marital Therapy, 22(4), 247–258.

233

Serwadda D., Mugerwa R. D., Sewankambo N. K., Lwegaba A., Carswell J. W., Kirya G. B., et al. (1985). Slim disease: A new disease in Uganda and its association with HTLV-III infection. Lancet, 19(2), 849–852.

Sherer, R. D., Jr, Bronson, J. D, Teter C. J, et al. (2004). Microeconomic loans and health education to families in impoverished communities: Implications for the HIV pandemic. Journal of the International Association of Physicians in AIDS Care, 3(4), 110–114.

Sherraden, M. (1986). School dropouts in perspective. Educational Forum, 51(1), 15–31.

Sherraden, M. (1990). Stakeholding: Notes in a theory of welfare based on assets. Social Service Review, 64(4), 580–601.

Sherraden, M. (1991). Assets and the Poor: A new American welfare policy. New York: M.E. Sharpe.

Slonim-Nevo, V., Auslander, W. F., Munro, J., & Ozawa, M. N. (1994). Knowledge and attitudes related to AIDS among African-American women: Results of an exploratory study. Ethnicity and Disease, 4, 68–76

Slonim-Nevo, V., Auslander, W.F., & Ozawa, M.N. (1995). Educational options and AIDS-related behaviors among troubled adolescents. Journal of Pediatric Psychology, 20, 41–60.

Ssewamala, F. M. (2005). Children Development Accounts in Africa: A pilot study. (Working paper No. 05-05). St. Louis: Washington University, Center for Social Development.

Ssewamala, F. M., & Curley, J. (2006). Asset ownership and school attendance of orphaned children in Sub-Saharan Africa. Social Development Issues, 28(2), 84–105.

Ssewamala, F. M., Alicea, S., Bannon, W. M., Jr., & Ismayilova, L. (2008). A novel economic intervention to reduce HIV risks among school-going AIDS orphans in rural Uganda. Journal of Adolescent Health, 42(1), 102–104.

Ssewamala, F. M., Bannon, W. M., Jr., Ismayilova, L., &. Alicea, S. (2008). Promoting health and educational opportunities for orphaned children through Children Development Accounts: Does gender matter? Manuscript submitted for publication.

Tangenberg, K. M. (2005). Faith-based human services initiatives: Considerations for social work practice and theory. Social Work, 50(3), 197–206.

Thiessen, C. Ssekubugu, R. Wagman, J. Kiddugavu, M. Wawer, M.J. Emanuel, E. Gray, R. Serwadda, D., & Grady, C. (2007). Personal and community benefits and harms of research: Views from Rakai, Uganda. AIDS, 21(18), 2493–2501.

Tierney, J. P., Grossman, J. B., Resch, N. L. (1995). Making a difference: An impact study of Big Brothers Big Sisters. Retrieved December 1, 2006, from http://www.ppv.org/ppv/publications/assets/111_publication.pdf

234

Tyler, T. (1997). Why do people rely upon others? Social identity and the social aspect of trust. Paper presented to the Russell-Sage workshop on trust. New York, April 26, 1997 (Unpublished).

United Nations Children’s Fund [UNICEF]. (2003). Africa’s orphaned generations. New York: UNICEF.

United Nations Children’s Fund [UNICEF]. (2008). State of the World’s Children 2008. New York: UNICEF.

United Nations Programme on HIV/AIDS [UNAIDS] (2006). Report on the Global AIDS Epidemic. Geneva: UNAIDS.

United Nations Programme on HIV/AIDS [UNAIDS], United Nations Children’s Fund [UNICEF], and United States Agency for International Development [USAID] (2004). Children on the brink: A joint report of new orphan estimates and a framework for action. New York: UNICEF/USAID.

World Bank. (2000a). Entering the 21st century: World development report 2000/2001. Attacking poverty, New York: Oxford University Press.

World Bank. (2000b). Helping the children: World Bank directory of HIV/AIDS interventions in Africa: Washington, DC: World Bank.

White, K. M., Terry, D. J., & Hogg, M. A. (1994). Safer sex behavior: The role of attitudes, norms, and control factors. Journal of Applied Social Psychology, 24(24), 2164–2192.

Zhan, M., & Sherraden, M. (2003). Assets, expectations, and children’s educational achievement in female-headed households. Social Service Review, 77(2)191–211.

235

EVALUATING THE POTENTIAL OF FAITH-BASED CORRECTIONAL MODELS: A CASE STUDY OF FLORIDA’S

FAITH-BASED AND CHARACTER-BASED INSTITUTIONS

Diana Brazzell Nancy La Vigne

As the popularity of faith-based programs has grown across the country, the criminal justice field has witnessed the development of numerous faith-based initiatives aimed at rehabilitating and supporting incarcerated people and those returning from prison and jail. The corrections system in the United States has increasingly welcomed partnerships with faith-based and community-based organizations, and churches and other religious institutions have formalized their longstanding role as a major source of community support for returning prisoners (Bright & Graham, 2007; Henriques & Lehren, 2006; Hercik et al., 2005; O’Connor, 2004, 2005). Embracing this trend, in 2003, Florida became the first state in the country to dedicate an entire publicly run correctional facility to a faith-based model. Currently, Florida operates three Faith- and Character-Based Institutions (FCBIs): Lawtey, a medium- and minimum-security male facility that houses 815 inmates; Hillsborough, a facility that houses 271 female inmates of all security classifications; and Wakulla, a maximum-security facility housing 1,741 male inmates (Florida Department of Corrections [FDOC], 2008b).

This paper presents the findings from a process and outcome evaluation of the FCBI initiative at the Lawtey and Hillsborough correctional facilities.1 Using both quantitative and qualitative data, the study explores three primary research questions: (1) What is the FCBI mission and program model? (2) How is this model being implemented? and (3) What are the outcomes of the FCBI program, particularly with respect to reincarceration? An earlier report focused on the program model and implementation issues (La Vigne, Brazzell, & Small, 2007). The present study uses newly available quantitative data to examine program outcomes, incorporating the previous qualitative findings to provide context regarding the FCBI program and its operations. The study’s findings are based on an analysis of FDOC inmate data; interviews with FCBI management, staff, and volunteers; focus groups with inmates participating in the FCBIs; and firsthand program observation.2

Given the current interest in faith-based corrections programs and the limited amount of research that exists on the impact such programs have on participants’ postrelease outcomes, this study offers valuable findings for researchers, policy makers, and corrections practitioners interested in faith-based models. To frame the discussion, the paper begins by exploring the existing research

237

on faith-based corrections programs, and then describes the FCBI initiative, its mission, and the program model. The paper presents quantitative findings on FCBI participant characteristics and recidivism outcomes, and concludes with a synthesis of the findings and a discussion of the implications for the criminal justice field.

RESEARCH CONTEXT

Religion and religious institutions have historically played an important role in the American penal system, beginning with the Protestant Christian influences on the development of the country’s first prisons (O’Connor, 2004). Even as prisons secularized in the latter half of the nineteenth century, churches and other religious institutions continued to send representatives into correctional institutions and to serve as a nexus of community support for those exiting prison. The past decade has witnessed a renewed emphasis on the role of religion in offender rehabilitation and increased visibility and funding for formal faith-based programs serving currently and formerly incarcerated people (Hercik et al. 2005; Mears et al. 2006; O’Connor, Duncan, & Quillard, 2006; White House, 2008). This trend is part of a broader growth in the popularity of faith-based models for addressing a range of social ills, an approach championed by President George W. Bush through the creation of the White House Office of Faith-Based and Community Initiatives (OFBCI) (McDaniel, Davis, & Neff, 2005; White House, 2008).

An important part of the expansion of faith-based responses to criminal justice problems has been the establishment of a number of comprehensive, faith-based prison programs (Bright & Graham 2007; Henriques & Lehren 2006). A 2005 review found that 21 state correctional systems and the federal prison system were operating faith-based residential programs or were in the process of developing them (National Institute of Corrections Information Center [NICIC], 2005).3 Despite the popularity of such programs, very little research exists on the effectiveness of faith-based programs (residential or nonresidential) that serve adult inmates (O’Connor, 2005; Mears et al., 2006; Sumter, 2006). To date, many of the studies of faith-based corrections programs have produced inconclusive or tentative findings and have been characterized by major methodological limitations (Mears et al., 2006; O’Connor, 2005).

Only two studies have evaluated comprehensive, residential faith-based prison programs in the United States, and neither found lower recidivism rates among program participants when compared with similar inmates from the general population. Johnson and Larson (2003) found no difference in rearrest and reincarceration rates between participants in the InnerChange Freedom Initiative in Texas and a matched comparison group. The subset of participants who officially graduated from the program did fare better than the comparison inmates, but graduation required

238

holding a job and belonging to a church after release, criteria that likely had a significant impact on recidivism outcomes independent of program participation. Hercik’s 2004 study of the Kairos Horizon program in Florida’s Tomoka Correctional Institution did not find a lower rearrest rate among program participants when compared with a matched comparison group, although the study’s sample sizes were small and the methodology was not described in enough detail to gauge its validity.

Although research evaluating faith-based corrections programs is scarce, a fairly strong theoretical foundation exists for efforts that aim to rehabilitate offenders and reduce recidivism through religious engagement. A significant body of literature links individual faith and participation in organized religion to reduced involvement in criminal activity among the general population (Hercik et al., 2005; Sumter, 2000). Studies of prisoners have found that spirituality is related to better in-prison behavior and greater levels of adjustment to prison (Clear & Sumter, 2002; Kerley, Matthews, & Blanchard, 2005). Other research has shown a connection between religious involvement and positive outcomes for health and well-being, such as lower rates of substance abuse and depression, which might improve reentry success and curb recidivism among people leaving prison (Johnson, Tompkins, & Webb, 2002). These findings suggest that programs that aim to cultivate and support religious involvement among inmates, such as the FCBI initiative, could reduce future criminal behavior and lead to other positive outcomes for participants.

THE FCBI INITIATIVE

The Florida Department of Corrections officially launched the FCBI initiative on December 24, 2003, when it converted Lawtey Correctional Institution to an FCBI serving medium- and minimum-security male inmates. The following April, the FDOC opened an FCBI for female inmates of all security classifications at Hillsborough Correctional Institution.4 The purpose of the FCBI initiative, according to the FDOC, is to offer a wide range of religious- and character-focused programming to inmates interested in “personal growth and character development” (FDOC, 2004). In the four years since their inception, Lawtey FCBI has served 3,616 inmates and Hillsborough FCBI has served 1,291 inmates (FDOC, 2008b).

Lawtey currently houses 815 inmates from 30 different faiths and Hillsborough houses 271 inmates belonging to 22 faiths (FDOC, 2008b).5 Using private funding, faith-based and character-based programming is provided by community volunteers from a range of religious and secular backgrounds. The role of the FDOC prison staff is to facilitate the provision of programming and maintain each FCBI as a “forum” in which community volunteers can offer

239

programs (FDOC, 2004). The immediate aims of the FCBI initiative are to provide a wider range of both secular and religious programming for inmates and to improve inmate behavior and institutional security. In the long term, the initiative aims to rehabilitate participants, promote their successful reintegration into the community after release, and ultimately reduce recidivism (FDOC, 2004, 2008c).

Programming

At the center of the FCBI model are the faith-based and character-based programs designed, funded, and implemented by community volunteers with the support and guidance of each FCBI’s chaplaincy staff. Although similar faith-based and character-based programs exist at other Florida correctional facilities, the number, range, and depth of the programming at the FCBIs are far greater. The programs available at Lawtey and Hillsborough range from explicitly religious activities, such as worship services and scriptural study, to personal relationship building through mentoring and small group activities, to character development programs covering topics such as parenting and anger management. Some programs are explicitly religious, whereas others present program topics from a secular perspective, addressing religion only if it is raised by inmates. More information on the specific programs available at Lawtey and Hillsborough can be found in Appendix B.

Several faith-based and character-based programs are offered each day and evening and many inmates attend multiple programs a week. Programs are typically scheduled for the same time each week; some, such as religious services and mentoring, are ongoing, while others, particularly the life-skills and character-development classes, follow a curriculum that lasts for a set number of weeks. In addition, there are occasional holiday celebrations, weekend retreats, and special worship events. As in other Florida prisons, Lawtey and Hillsborough also offer educational opportunities, vocational training, wellness programs, and reentry programs funded and staffed by the FDOC.

An analysis of one month of data on participation in the faith-based and character-based activities found that program involvement varies from inmate to inmate in terms of both the quantity and types of programs in which an inmate participates.6 The set of available programs also varies across the FCBI facilities and over time, primarily because of the reliance on volunteers to provide programming. Because of this variation in program activity at the facility and individual levels, participation in an FCBI cannot be treated as a fixed experience that is identical for all inmates. The lack of standardized programming should be kept in mind when

240

considering measurements of program effectiveness and opportunities for replicating the FCBI model elsewhere.

Volunteers and Chaplains

The FCBIs offer an extensive range of religious and character development opportunities by leveraging private resources from the community, particularly human capital, at little cost to the state. At Lawtey, 46 community groups provide programming to 815 inmates and 57 community groups currently serve Hillsborough’s 271 inmates (FDOC, 2008b). Although volunteers from secular groups and several religious traditions are active in both facilities, the vast majority of volunteers are Christian and participate through their churches. As at other FDOC facilities, FCBI volunteers receive four hours of standardized training that focus on guidelines for working safely and effectively in a prison setting. Some volunteers receive additional training from their sponsoring organizations.

Volunteers are supported and guided in their work by the FCBI prison chaplains, who serve as de facto program coordinators. All Florida prisons have chaplains who work to ensure that the religious needs of inmates are met, but the chaplaincy position at the FCBIs involves a greatly expanded set of responsibilities. FCBI chaplains recruit, train, and support community volunteers; assist with program development; manage the program schedule and work to fill program gaps; and act as the liaison between volunteers and prison administrators. According to one FCBI chaplain, management skills are as important to the position as religious training.

Environment and Staff

The theory behind the FCBI initiative is not only to offer inmates more extensive faith-based and character-based programming, but to concentrate these programs in select facilities and to cultivate an atmosphere within these facilities that supports rehabilitation. The FCBIs attract inmates interested in positive change and house them together in an environment infused with programming and community volunteers. By saturating the facility with opportunities for rehabilitation, the FCBI model aims to develop a supportive environment within which programming can have a greater impact.

Many correctional staff view contributing to this positive facility environment as part of their job responsibilities and therefore they approach their jobs in a manner conducive to inmate rehabilitation. In practice, this rehabilitative correctional philosophy includes relating to inmates in a friendly and respectful manner; serving as a positive behavioral example for inmates; and, when appropriate, resolving problems with inmates informally through discussion and mediation rather than punitive responses. Many of the volunteers, chaplains, management, and staff

241

interviewed for this study emphasized the importance to the FCBI initiative of this shift in correctional culture, although not all FCBI staff have adopted a rehabilitative philosophy (for more on this point, see La Vigne et al., 2007).

Inmate Participants

Inmates volunteer for FCBI placement and may be of any religious faith or none at all.7 The findings from interviews with FCBI staff, chaplains, and inmates suggest that the majority of inmates who volunteer do so because they have a genuine interest in the FCBI program, although a number are driven by other interests, such as wanting to be closer to family or to be housed in a less harsh prison environment. Those inmates who express interest in being housed in an FCBI are placed on a waiting list, which currently comprises a few thousand inmates. On reaching the top of the list, inmates are screened for eligibility.

To be eligible for the FCBI, inmates must fit the inmate profile of the given FCBI facility (e.g., in terms of security level, offense type, time to release), must be in general population housing status and not on work release, and cannot have received a disciplinary report resulting in confinement in the previous 90 days. At Lawtey and Hillsborough, the first criterion generally excludes inmates who are sex offenders or have more than five years remaining on their sentence. If determined to be eligible for placement, inmates are transferred to the appropriate FCBI facility. Inmates must meet minimum behavioral standards and participate in at least one program session a week to remain at the FCBI.8 A wide range of secular and religious options can satisfy the weekly program participation requirement, including faith-based and character-based programs, educational classes, and vocational training.9 Unless an inmate requests to leave or is transferred out for failing to meet the behavioral and program participation standards, he or she will typically remain in the FCBI until release. In addition, an inmate can choose to be transferred out at any time by submitting a request to his or her classification officer.

FCBI OUTCOMES

To better understand the characteristics, experiences, and outcomes of FCBI participants, FDOC administrative data on FCBI inmates and inmates in Florida’s general prison population were analyzed. The dataset provided information on all individuals incarcerated in the Florida prison system on September 30, 2004, including demographic characteristics, criminal history, in-prison behavior, FCBI stays, and releases and returns to prison through February 8, 2008. The following section presents findings from analyses of these data, supplemented with qualitative perspectives on program outcomes. Because of differences between Lawtey and Hillsborough identified in other parts of the study, all analyses were conducted separately for males and females.10

242

Characteristics of FCBI Inmates

The characteristics of inmates housed in the Lawtey and Hillsborough FCBIs on September 30, 2004 (n=696 males, 261 females) were analyzed and compared with the individuals incarcerated in all other Florida prisons on that date (n=74,006 males, 4,802 females). Tables showing the full results of the chi-square tests and independent sample t-tests are included in Appendix C. Given that inmates self-select into the FCBIs and that they must meet specific eligibility criteria to enter and remain in the facilities, it is not surprising that some significant differences existed between the FCBI participants and the general Florida prison population.

While male FCBI inmates were demographically similar to the male general prison population in terms of age and race, they had significantly different criminal backgrounds. Only 29% of Lawtey inmates were incarcerated for a violent offense compared with 54% of the male general population. There were no sex offenders at Lawtey and a much smaller number of murder/manslaughter offenders than in the general population; the lower numbers of these types of offenders were balanced by a greater number of drug offenders. On average, Lawtey inmates had shorter sentences, had spent less time in prison during their current incarceration, and had less time remaining until release than general population male inmates. These differences in criminal history can be explained primarily by the fact that only minimum- and medium-security inmates with less than five years to release are eligible to be housed in Lawtey, but differences may also result from other program eligibility criteria (e.g., in-prison behavior) or inmate self-selection into the program. Interestingly, however, Lawtey inmates had more prior incarcerations than the typical male inmate housed in the general population.

In contrast to the males, the criminal histories of female FCBI inmates were not significantly different from those of the female general prison population on most measures, probably because, unlike Lawtey, Hillsborough houses inmates of all security classifications. FCBI and general population female inmates were comparable in terms of the proportion serving time for violent offenses and drug offenses, the number of prior incarcerations, the length of time served, and the expected sentence length. However, no female FCBI inmates were sex offenders, fewer were incarcerated for murder/manslaughter offenses than in the female general population, and the FCBI inmates had a shorter time remaining until release than the average female offender. The demographic characteristics of female FCBI inmates were slightly different from the general female prison population in that the FCBI inmates tended to be older and were more likely to be White.

243

The differences between the FCBI inmates and the general prison population outlined above must be taken into account when considering the applicability of any findings on program effectiveness for the general prison population. For example, because sex offenders are not eligible to be housed at Lawtey and Hillsborough, there is no way of knowing how FCBI participation might impact sex offenders. In addition, there may be other differences between the FCBI inmates and the general prison population that were not possible to measure using the available data for this study. The dataset did not capture information on the inmates’ attitudes and beliefs, the intensity and nature of their religious involvement, or their motivation for rehabilitation and reentry success.

The FCBI Experience

Examining the program experiences of the FCBI inmates in the sample (n=696 males, 261 females) showed that most inmates spent no more than a year and a half to two years in the FCBI. Among male FCBI inmates, 94% spent at least six months at Lawtey and 71% spent at least a year there.11 Among female FBCI inmates, 81% spent at least six months at Hillsborough and 50% spent at least a year there. Yet less than one out of five inmates (18% at Lawtey, 16% at Hillsborough) who were in an FCBI on September 30, 2004, were still housed there on May 31, 2006, 20 months later. Of those who left the FCBIs, 62% from Lawtey and 67% from Hillsborough had been released; the remainder had been transferred to other facilities.12

Other than the length of time spent in the FCBI, there is no specific information on the program participation of inmates in the sample during their time in the facilities, such as the number and types of programs in which they participated and their involvement in secular versus faith-based programming.13 This study, therefore, is limited to an exploration of the overall impact of being housed in an FCBI, a valuable question given that the model’s effects are intended to stem from both program participation and the overall environment in the FCBI facilities. Such an approach is appropriate for understanding the FCBI model as it is currently implemented, without a standardized curriculum or a requirement that inmates participate in specific programs. If more detailed individual-level program participation data were available, a subsequent evaluation could examine how different levels and types of program involvement influence outcomes.

Reincarceration Outcomes

To examine the hypothesis that the FCBI model reduces recidivism, the reincarceration rates of FCBI inmates were compared with those of a matched group of general population inmates who did not participate in the FCBI. The treatment group comprised inmates housed in Lawtey or Hillsborough on September 30, 2004, who were released at some point between November 1,

244

2004, and November 31, 2005, and who spent at least 3 months in the FCBI.14 On average, the male inmates in the treatment group were housed at Lawtey for 13 months and the female inmates spent 9 months in Hillsborough.15 These FCBI inmates were matched to similar inmates who were incarcerated in any of Florida’s other state prisons on September 30, 2004; were released between November 1, 2004, and November 30, 2005; and were not housed in the FCBI at any point in time.

Using a one-to-one categorical exact matching technique, inmates from the general prison population group were matched to the inmates from the FCBI group on the following factors: gender, age, race, primary offense type, violent or nonviolent offense, number of prior incarcerations, time served, and disciplinary report rate (a measure of in-prison behavior prior to program participation).16 This method matched each FCBI inmate with one non-FCBI inmate using the factors listed above and produced treatment and comparison groups for males and females that were statistically similar in terms of the factors used to match the groups.17 Because of the limitations of the dataset, it was not possible to account for other, less tangible factors (e.g., religiosity or personal motivation) that might vary between inmates who did and did not participate in the FCBI program. The differences between the treatment and comparison groups on these factors are unknown, although it would be reasonable to hypothesize that some inmates who volunteer for the FCBI are more religious or more motivated than the average Florida inmate.

The outcome measure used in this analysis is reincarceration in the Florida state prison system within 26 months after release.18 As with the rest of the analysis, outcomes for males (n=189 treatment and 189 comparison) and females (n=100 treatment and 100 comparison) were analyzed separately. For both males and females, no statistically significant difference was found in the proportion of FCBI and non-FCBI inmates returned to prison within 12, 18, 24, and 26 months of release.19 Table 1 presents the reincarceration rates for the treatment and comparison groups, and more detailed tables in Appendix C provide chi-square values. There was also no statistically significant difference between the treatment and comparison groups in the proportion of inmates who were reincarcerated for a technical violation (i.e., a parole or probation violation) versus a new crime.20

245

Table 1. Reincarceration Outcomes of FCBI Inmates and a Matched Comparison Group

Males Females

Returned to prison within

FCBI Inmates (n=189)

Comparison Group (n=189)

FCBI Inmates (n=100)

Comparison Group (n=100)

12 months 8 4.2%

8 4.2%

4 4.0%

7 7.0%

18 months 20 10.6%

19 10.1%

9 9.0%

8 8.0%

24 months 23 12.2%

32 16.9%

14 14.0%

11 11.0%

26 months 27 14.3%

34 18.0%

15 15.0%

12 12.0%

These findings suggest that the FCBI program does not produce a statistically significant reduction in recidivism for either males or females in the first 2 years out of prison. Because the sample sizes are relatively small, however, the possibility of a small or moderate effect on recidivism cannot be ruled out. Yet it should be noted that for females, the effect, while not statistically significant, was in the opposite direction from what was expected: a greater proportion of female FCBI inmates were reincarcerated at 18, 24, and 26 months than the inmates from the comparison group.

The impact of FCBI participation on the quality of life of inmates who were not reincarcerated is unknown, and while the initiative may not be achieving its goal of significantly reducing recidivism, it may be achieving other program goals related to inmate rehabilitation. This analysis does not examine how former participants fared in terms of employment, family relationships, drug abuse, and additional reentry outcomes other than reincarceration, nor does it explore the impact of FCBI participation on in-prison behavior. Initially, a comparison of disciplinary infractions among FCBI inmates and their general population counterparts was to be included. However, interviews with FCBI staff and management suggested that disciplinary reports are employed differently in the FCBIs than in other FDOC facilities.21 These findings raised concerns about the reliability of comparing disciplinary infractions across facilities.

Some additional methodological limitations concerning this analysis were mentioned earlier, but they merit repeating. First, while all inmates in the treatment group were in an FCBI for at least 3 months, no other measure of treatment exposure or intensity was available. Future analyses that

246

take into account the number and types of faith-based and character-based programs in which inmates participate might find different outcomes for FCBI inmates with different levels of program participation. Second, in the current analysis, it is possible that the FCBI inmates differ from the comparison group in certain ways that cannot be measured. FCBI participants may already be more likely to succeed after release, regardless of program participation because they are more motivated, have better attitudes, or are more religious. The finding that the FCBI program has no effect, however, suggests that whatever selection bias is taking place is not significant enough to create measurable differences between the groups in terms of reincarceration outcomes.

Although the quantitative analysis shows no evidence of an effect on recidivism from FCBI participation, anecdotal reports from interviews with correctional staff, facility management, chaplains, and volunteers suggest that the FCBI program is producing positive outcomes during incarceration and possibly after release. These individuals reported observing attitudinal and behavioral improvements among inmate participants, including positive changes in inmates’ appearances, behaviors, language, and attitudes; improvements in the way inmates interact with one another and with staff and volunteers; and an increased sense of hope, responsibility, and self-worth among participants. The general opinion of those interviewed was that these positive outcomes are indicative of potential long-term changes that will result in greater success on release from prison. Although the FCBIs do not formally track inmates after release, the chaplains and some correctional staff do keep in contact with certain former inmates. Many suggested that, as a result of FCBI participation, these inmates had a smoother transition back into the community and a stronger foundation for building a successful, crime-free life on the outside. Some FCBI staff, however, believed that the effect was one of self-selection, suggesting that participants who do not recidivate are those who were likely to succeed regardless of FCBI participation.

In addition to improving outcomes for individual inmates, the FCBI program aims to produce a safer, more positive facility environment. Most management and correctional staff interviewed felt that the working environment in the FCBIs is safer and more pleasant than that of traditional correctional facilities. They reported that inmate conflicts are less frequent, relations between inmates and staff are better, inmates are more respectful toward staff and volunteers, inmates keep the facility cleaner, and incidents of misbehavior are more likely to be isolated occurrences rather than ongoing problems. Volunteers generally found the FCBI environment to be welcoming and felt that a positive tone is fostered by both staff and inmates. The inmates who

247

participated in the focus groups were also strong proponents of the FCBI environment, finding it safer, less stressful, and more conducive to positive change than traditional prisons.

DISCUSSION AND IMPLICATIONS FOR THE FIELD

The quantitative analysis provides no evidence that the FCBI program is reducing recidivism, as measured by reincarceration, for either males or females in the first 2 years out of prison. The effect on other pre- and postrelease outcomes remains unknown, as these were not examined in the current study. The analysis is not extensive enough to pronounce the model a failure, especially in light of anecdotal evidence of program success. However, the findings should give pause to policy makers considering faith-based solutions to criminal justice problems. Despite that fact that at the moment faith-based correctional models are receiving a great deal of attention and support among policy makers and positive reviews from providers and participants, little evidence exists to support their effectiveness. The current study can help inform the development, refinement, and future evaluation of other faith-based models, for although the FCBI initiative lacks demonstrated success in reducing recidivism, it does hold valuable lessons for other faith-based corrections initiatives.

Developing Evidence-Based Practice

To support more rigorous testing of faith-based corrections models, policy makers and corrections officials implementing faith-based initiatives need to clearly define the intended outcomes of the initiative, create benchmarks for measuring success, and implement systems for tracking and analyzing outcomes data. To facilitate evaluation, program offerings should be standardized to provide a measurable treatment that is consistent over time and across facilities. When developing programming, secular models that have been shown to be effective should be reviewed and possibly adapted to the faith-based environment.

Leveraging Private Resources

One of the greatest benefits of the FCBI model is that it brings extensive program offerings into facilities at a minimal cost. The primary costs of the FCBI initiative are the expenses for a somewhat expanded chaplaincy team and the additional demands on staff time driven by increased volunteer and inmate activity, although these added demands may be balanced with reductions in behavioral problems among FCBI inmates.22 The FCBIs demonstrate that local volunteers, particularly from the faith community, are a tremendous untapped resource for both faith-based and traditional prisons and a potential asset for budget-strapped systems.

248

Extending Success Beyond Release

Like all corrections programs, faith-based initiatives intended to have a long-term impact need to address the reentry process. In this study, the anecdotal evidence suggests positive outcomes inside the FCBI facilities, but the quantitative analysis does not demonstrate increased success after release. One possible explanation for these findings is that the FCBI initiative is not providing the continuity of support needed to extend program benefits beyond the prison walls. Indeed, one of the findings of the earlier process evaluation was that the reentry components of the FCBI model are underdeveloped. Corrections officials designing faith-based initiatives should incorporate mechanisms for extending the support and guidance provided by the programming through release and reentry. They should also ensure that the faith-based and character-development activities are complemented with vocational training and educational opportunities that develop the concrete skills necessary for reentry success.

Avoiding Religious Discrimination

The FCBI experience has demonstrated that faith-based efforts in public correctional facilities can successfully serve inmates from different faiths and avoid religious discrimination. Many of the most popular faith-based corrections programs, such as the InnerChange Freedom Initiative, are exclusively Christian and explicitly religious, and these types of programs have run into legal challenges because of their bias toward a specific creed or denomination and the blurring of lines between church and state (Henriques & Lehren, 2006). From all indications, the FDOC carefully designed the FCBI program to draw on the perceived benefits of a faith-based approach while avoiding some of the church-state conflicts encountered by other faith-based corrections programs. It did so by having all faith-based programming provided by volunteers using private funds, including a nonreligious character-development component in the model, making participation voluntary, explicitly avoiding any form of religious test for program participation, and recruiting volunteers from both Christian and non-Christian faiths (La Vigne et al., 2007). The FCBI model presents a promising approach to issues of religious diversity that could benefit other faith-based corrections initiatives.

CONCLUSION

Despite the lack of quantitative findings demonstrating any positive effect on recidivism from the FCBI initiative, elements of the model hold value for both faith-based and traditional correctional facilities. The lessons outlined above and others discussed in an earlier report (La Vigne et al., 2007) may prove beneficial to policy makers and corrections officials considering faith-based models. Researchers and decision makers should continue to evaluate faith-based

249

corrections models, and in particular devote time to clearly defining and tracking outcomes and quantifying program costs. As for the FCBI initiative and others like it, standardized programming that draws in part on proven rehabilitative models could considerably improve program effectiveness while retaining many of the desirable elements of the existing model.

ACKNOWLEDGMENTS

The authors thank the Annie E. Casey Foundation, which provided funding for this study, as well as the Florida Department of Corrections and the many FDOC employees who shared their expertise and knowledge. In particular, Dr. David Ensley, Dr. John L. Lewis, and Dr. Frank Liu from the FDOC Bureau of Research and Data Analysis provided data and guidance for the quantitative analysis. The authors also thank their institutional home, The Urban Institute (UI), and their peers from UI and RTI International who reviewed drafts of the paper.

NOTES

1. Wakulla was converted to an FCBI on March 1, 2006. Because Wakulla has been operating as an FCBI for a shorter period than the other two facilities, it is not included in this study.

2. See Appendix A for a detailed discussion of the qualitative data used in this study. The qualitative data were collected during site visits to Lawtey and Hillsborough in summer 2006, when the FCBI programs at each facility had been operating for over two years.

3. The NICIC review only included programs where, as in the FCBIs, participating inmates were housed separately from the general prison population.

4. As discussed in Note 1, Florida’s Wakulla Correctional Institution is also an FCBI but was not included as part of this study.

5. Although participating inmates are primarily Christian, 10% are of non-Christian faiths and 11% are not religious or do not identify with a specific religious affiliation (FDOC, 2008b). The breakdown of religious affiliations among FCBI inmates is fairly similar to the breakdown for the overall Florida prison population (FDOC, 2008a, 2008b).

6. Program participation data provided by the chaplains from October 2006 at Lawtey (n=694 inmates) and from June 2006 at Hillsborough (n=348 inmates) were analyzed.

7. FDOC policy explicitly states that “an inmate’s religious faith or lack thereof [is] not… considered” in determining program eligibility (FDOC, 2004).

8. Inmates will be transferred out of the FCBI if they receive a disciplinary report resulting in confinement or an excessive number of overall disciplinary reports. Decisions on expulsion are made on a case-by-case basis and take into account both the number and types of disciplinary reports.

9. Certain faith-based activities, such as religious education classes and small group studies, count toward the program requirement, but weekly religious services, holidays, and special events do not.

250

10. The qualitative portion of this study indicates that the male experience at Lawtey and the female experience at Hillsborough are different enough, in terms of both environment and program offerings, that the two FCBIs should be conceptualized as distinct experiences with potentially differing effects. Prior research also indicates that male and female inmates typically have different characteristics and distinct prison experiences (Bloom, Owen, & Covington, 2003).

11. Some inmates were incarcerated in Lawtey before it was converted to an FCBI and chose to remain there after the conversion. The figures cited here only include time spent in Lawtey when it was an FCBI.

12. It is not possible to know from the dataset whether inmates who were transferred from the FCBI to another prison were moved because they requested to the leave the FCBI, because they did not meet the eligibility requirements to remain in the program, or for some other reason, such as being granted work release.

13. Although the FCBI chaplains do track inmate participation in four broad categories of faith-based and character-based programming, their records do not clearly distinguish the amounts and specific types of programming. This precluded quantifying the faith-based, character-based, and secular (vocational, educational, etc.) program participation of inmates in the sample.

14. Inmates in the treatment group could have entered the FCBI facility at any point in time before September 30, 2004, from several months prior to one day prior.

15. The figures cited here only include time spent in Lawtey and Hillsborough when they were FCBIs.

16. A report by the FDOC on recidivism among Florida inmates identifies several of the factors used in this study to match the samples as important predictors of reincarceration, particularly age and number of prior incarcerations (FDOC, 2003).

17. Tables showing the characteristics of the treatment and comparison groups are included in Appendix C. The only significant distinction between the groups is that female FCBI inmates served slightly longer sentences than the comparison females, a difference that was statistically significant at the 0.10 level.

18. Although only reincarceration in the state of Florida was analyzed, there is no reason to believe that FCBI inmates are more or less likely than non-FCBI inmates to be reincarcerated in another state as compared with Florida.

19. Outcomes up to 26 months after release were analyzed because all inmates were released from prison at least 26 months prior to the last day for which there are reincarceration data. For time periods beyond 26 months, reincarceration data were not available for all inmates in the treatment and comparison groups.

20. Among male FCBI inmates who were reincarcerated, 44.1% were returned for a technical violation, compared with 39.1% of male inmates in the matched comparison group who were reincarcerated. Among female inmates who were reincarcerated, 62.5% in the FCBI group were reincarcerated for a technical violation, compared with 53.3% in the non-FCBI group. Neither of these differences was statistically significant at the 0.10 level.

251

21. Some respondents indicated that FCBI inmates are held to higher behavioral standards and written up for more minor offenses than at other FDOC facilities, whereas other respondents reported that officers in the FCBIs are more likely to resolve problems through verbal warnings or other approaches that do not result in a formal disciplinary report.

22. No attempt was made to quantify these costs for the current study.

APPENDIX A: QUALITATIVE RESEARCH METHODOLOGY

Qualitative data for this study were collected in summer 2006 during site visits to the Hillsborough and Lawtey FCBIs. Interviews were conducted with FCBI staff, management, and volunteers as well as focus groups with FCBI inmates. In addition, the facilities were toured and a handful of faith-based and character-based programs at each facility were observed. These research activities were supplemented with documents and data obtained from the FDOC and information from e-mail and telephone communications with FDOC officials.

To obtain the perspectives of FCBI staff, management, and volunteers, semistructured, one-on-one interviews were conducted that lasted 30 minutes to an hour. At Hillsborough, 22 individuals were interviewed: one upper management staff, two correctional officers, one chaplaincy staff member, four civilian support staff (e.g., medical, mental health, classification), and 14 community volunteers. At Lawtey, 14 individuals were interviewed: two upper management staff, four correctional officers, two chaplaincy staff members, one civilian support staff, and five volunteers.

One 90-minute focus group was conducted at each facility with randomly selected inmates. The Hillsborough focus group included seven participants from Protestant and Catholic Christian faiths. Three participants are Black, three are White, and one is Hispanic. The average age of the participants was 40. At Lawtey, the focus group included six inmates from Protestant, Catholic, and Jewish faiths. Two participants are black, two are White, one is Hispanic, and one is multiracial. The average age of the participants was 33.

APPENDIX B: FCBI PROGRAMMING

As mentioned earlier, the range of programming available in the FCBIs varies over time, and each inmate participates in a different set of programs during his or her stay. Nonetheless, a review of the types of programs offered can provide a sense of the scope of activities in which FCBI inmates are involved. Tables B-1 and B-2 provide a snapshot of the programs available at Lawtey and Hillsborough, respectively. The tables divide programming into three categories—

252

life skills, character building and wellness, and religious programs—a rubric developed (by the authors) for the purpose of more clearly categorizing different program offerings.

Some life skills and wellness programs are funded by the FDOC and taught by FDOC staff or private contractors. Other programs that fall into these categories are provided by community volunteers. Religious programming is offered either by community volunteers or the facility chaplains. Some programs follow established program models or are offered by individuals with expertise in a given field, such as an anger management program led by a volunteer who is a licensed psychologist. Other programs are developed by volunteers, in many cases based on their personal experiences and religious beliefs. The formal qualifications vary among volunteers leading programs. Religious education classes, for example, may be led by a local pastor or other religious leader, a lay member of a religious institution, or even an inmate.

Many of the types of programs listed in Tables B-1 and B-2 are available in other Florida correctional facilities, but the number and range of the religious and character-building programs is much greater in the FCBIs. Certain programs, such as adult basic education, the Modality substance abuse treatment program, and the mandatory reentry transition program, are offered in all FDOC facilities. The availability of most other programming, including offerings such as vocational training, varies from facility to facility.

One notable program offered in both of these FCBI facilities is the mentoring program, in which community members are matched with inmates in an ongoing faith-based mentoring relationship. Mentoring was envisioned as a major component of the FCBI initiative and is explicitly mentioned in the official FDOC policy outlining the FCBI model (FDOC, 2004). During site visits in summer 2006, Hillsborough had 85 mentor-mentee pairings. Lawtey had 150 mentors, but because the demand for mentors exceeded their availability, some Lawtey mentors were working with inmates in small groups rather than one-on-one. Inmates request to participate in the program, and the chaplains pair inmates with mentors who they believe provide a good match. Mentors meet with their mentees weekly, and the relationship may last as long as the individual is incarcerated. Although many mentor–mentee relationships involve a faith element, the vast majority of mentors said their focus was on personal relationship building and individualized support rather than religion.

253

Table B-1. Programs at Lawtey FCBI

Life Skills Character Building and Wellness Religious Programming Education

Adult Basic Education Program Correspondence Courses General Education Degree (GED)

Program Literacy Program

Vocational Carpentry Class Computer Class Drafting Class Prison Industries Garment

Program

Reentry FDOC Transition Class

(mandatory) Faith-based classes Financial Peace GED Tutoring Preparing for Employment Toastmasters (public speaking) WorkNet

Libraries Institutional Library Law Library

Substance abuse treatment Alcoholics Anonymous Modality Program (official FDOC

substance abuse treatment program)

Narcotics Anonymous Overcoming Addictions (faith-

based)

Wellness programs Smoking Cessation Wellness Education

Arts Choir Drama

Faith-based classes Anger Management Bridge Builders (overcoming

addictions of all types) Laughter From Purity (exiting

same-sex relationships) Manhood Series Parenting Class Practical Christian Living Taking Responsibility

Other Seeing-Eye Dog Training Class

Christian religious services Baptist Church Crusades Christian Praise/Worship (services

offered by 6 ministries) Hispanic Christian Services Motorcycle Ministry Roman Catholic Mass

Other religious services Inter-Faith Meditation Group Islamic Jumah Prayer Jewish Sabbath Prayer Service Native American Sacred Pipe

Service

Small group religious study Catholic Fellowship Christian Bible Study Friendly Faith Sharing (Christian) Hebrew-Israelite Bible Study Hispanic Christian Bible study Islamic Taleem Studies Jehovah Witness Bible Study &

Prayer Jewish Torah Studies Native American Spirituality Class Roman Catholic Catechism

Studies Scientology Studies

Libraries Audio-Visual Library Chapel Library

Other Holy Day Observances (All

Faiths) Faith-based Mentoring Program

Source: This table is based on a program list provided by the Lawtey Chaplain’s Office on February 1, 2007, as well as information from the FDOC Web site (FDOC, 2007b).

254

Table B-2. Programs at Hillsborough FCBI

Life Skills Character Building and Wellness Religious Programming Education

Adult Basic Education Program General Education Degree (GED)

Program Literacy Program Correspondence Courses

Vocational Carpentry Class Cabinetmaking Class Commercial Foods/Culinary Arts

Reentry FDOC Transition Class

(mandatory)

Faith-based classes Finances Life Skills

Libraries Institutional Library Law Library

Other Junior Achievement Head Start for Success PROP

Substance abuse treatment Alcoholics Anonymous Modality Program (official

FDOC substance abuse treatment program)

Wellness programs ESUBA (program for victims of

abuse) Wellness & Health Betterment Women’s Health

Arts Art Choir Creative Writing Voice Training

Other Anger Management Celebrate Recovery Etiquette Grief and Loss Healthy Relationships Parenting Root Issues Second Chances (self

transformation) Self Esteem Spiritual Journaling Story Time Moms Tackling Tough Skills

Christian religious services Christian Ministries (services

offered by 20 ministries) Frontline Film Fest Ministry Full Gospel Business Men Hispanic Jail Ministry Kairos International Praise in Motion Dancers Roman Catholic Mass

Other religious services/study Buddhist Instruction (as needed) Christian Bible Study (services

offered by 3 ministries) Hispanic Bible Study Islamic Jumah Prayer Islamic Taleem Studies Jehovah’s Witness Bible

Study/Prayer Purpose Driven Life Roman Catholic Bible Study Wiccan Instruction (as needed)

Libraries Audio-Visual Library Chapel Library

Other Faith-based Mentoring Program

Source: This table is based on a program list provided by the Hillsborough Chaplain’s Office on June 22, 2006, as well as information from the FDOC Web site (FDOC, 2007a).

255

APPENDIX C: STATISTICAL FINDINGS

Characteristics of FCBI Inmates

Inmates housed in Lawtey and Hillsborough FCBIs on September 30, 2004 (n=696 males, 261 females) were compared with all other individuals incarcerated in Florida prisons on that date (n=74,006 males, 4,802 females)—except the 2,023 male inmates and 284 female inmates on the FCBI waitlist, who were excluded from the analysis. Tables C-1 and C-2 present the characteristics of the FCBI and general population inmates.

Table C-1. FCBI and General Population Inmate Characteristics

Proportion of Males Proportion of Females

Variable FCBI

Inmates General

Population Difference FCBI

Inmates General

Population Difference

Race

% White 42.0 44.9 2.9 59.4 51.8 7.6 **

% African American 55.0 52.2 2.8 37.9 42.9 5.0

% Hispanic 2.6 2.5 0.1 2.7 4.8 2.1

Primary Offense

% Violent 28.6 53.6 25.0 ** 37.2 34.5 2.7

% Murder/Manslaughter 4.3 13.7 9.4 ** 7.3 12.4 5.1 **

% Sex Crime 0 12.2 12.2 ** 0 1.6 1.6 **

% Drug Crime 32.0 18.7 13.3 ** 27.2 29.4 2.2

256

Table C-2. FCBI and General Population Inmate Characteristics

Males – Mean Values Females – Mean Values

Variable FCBI

Inmates General

Population Difference FCBI

Inmates General

Population Difference

Age (years) 36.5 36.0 0.5 37.6 35.8 1.8 **

Number of prior incarcerations

1.27 1.09 0.19 ** 0.72 0.65 0.07

Time in prison (months) 34.1 56.7 22.6 ** 25.2 27.0 1.8

Time until expected release (months)

24.4 57.5 33.1 ** 22.1 29.6 7.5 **

Sentence (months) 58.5 99.8 41.3 ** 47.3 51.1 3.8

Notes: All data reflect inmate characteristics as of September 30, 2004. There are nine primary offense categories: murder/manslaughter, sex crimes, drug crimes, robbery, burglary, property theft/fraud/damage, violent – other, weapons, and other. The four categories included in Table C-1 are those for which the greatest differences between the groups were observed. Sentences were calculated as the difference between the admission date and the expected release date.

* indicates differences that are statistically significant at p≤0.10. ** indicates differences that are statistically significant at p≤0.05.

Treatment and Comparison Groups

To create a comparison group, FCBI inmates were matched to inmates from the general prison population on the following factors: gender, age, race, primary offense type, violent or nonviolent offense, number of prior incarcerations, time served, and disciplinary report rate. As shown in Tables C-3 and C-4, there was only one statistically significant difference (at the 0.10 level) between the treatment and comparison groups on the factors used to match the groups.

257

Table C-3. Treatment and Comparison Group Characteristics

Proportion of Males Proportion of Females

Variable FCBI

Inmates General

Population Difference FCBI

Inmates General

Population Difference

Race

% White 38.1 38.1 0 58.0 58.0 0

% African American 59.8 59.8 0 39.0 39.0 0

% Hispanic 2.1 2.1 0 3.0 3.0 0

Primary Offense

% Violent 28.6 28.6 0 31.0 31.0 0

% Murder/Manslaughter 2.6 2.6 0 3.0 3.0 0

% Sex Crime 0 0 0 0 0 0

% Drug Crime 30.7 30.7 0 24.0 24.0 0

Table C-4. Treatment and Comparison Group Characteristics

Males – Mean Values Females – Mean Values

Variable FCBI

Inmates General

Population Difference FCBI

Inmates General

Population Difference

Age (years) 37.5 37.3 0.1 38.3 37.3 1.0

Number of prior incarcerations

1.25 1.23 0.03 0.73 0.70 0.03

Time served (months) 42.8 39.9 2.9 33.6 27.7 5.9 *

In-prison disciplinary rate (reports per month)

0.030 0.022 0.008 0.078 0.069 0.009

Notes. All data reflect inmate characteristics as of September 30, 2004, except time served, which is the actual time served from the beginning of the incarceration through release. See the notes on the previous page for a description of the primary offense categories. The disciplinary rate measures an inmate’s in-prison behavior prior to receiving treatment and is calculated as the number of disciplinary reports received per month during the current incarceration prior to entering the FCBI (for FCBI inmates) or prior to the study start date of September 30, 2004 (for non-FCBI inmates).

* indicates differences that are statistically significant at p≤0.10.

** indicates differences that are statistically significant at p≤0.05.

258

Recidivism Outcomes

The reincarceration outcomes of the treatment and comparison groups were analyzed at 12, 18, 24, and 26 months after release from prison. Tables C-5 and C-6 present findings from the chi-square tests for males and females, respectively. No statistically significant differences were found between the groups.

Table C-5. Reincarceration Outcomes of Male FCBI Inmates and a Matched Comparison Group

Returned to prison within

FCBI Inmates (n=189)

Comparison Group (n=189) Chi-square Value

Statistical Significance

12 months 8 4.2%

8 4.2%

0.000 1.000

18 months 20 10.6%

19 10.1%

0.029 0.866

24 months 23 12.2%

32 16.9%

1.724 0.189

26 months 27 14.3%

34 18.0%

0.958 0.328

Table C-6 Reincarceration Outcomes of Female FCBI Inmates and a Matched Comparison Group

Returned to prison within

FCBI Inmates (n=189)

Comparison Group (n=189) Chi-square Value

Statistical Significance

12 months 4 4.0%

7 7.0%

0.866 0.352

18 months 9 9.0%

8 8.0%

0.064 0.800

24 months 14 14.0%

11 11.0%

0.411 0.521

26 months 15 15.0%

12 12.0%

0.385 0.535

259

REFERENCES

Bloom, B., Owen, B., & Covington, S. (2003). Gender-responsive strategies: Research, practice, and guiding principles for women offenders (NIC 018017). Washington, DC: National Institute of Corrections.

Bright, L. C., & Graham, M. G. (2007). Faith-based programs give facilities a helping hand (NCJRS 221896). Washington, DC: U.S. Department of Justice.

Clear, T. R., & Sumter, M. T. (2002). Prisoners, prison and religion: Religion and adjustment to prison. Journal of Offender Rehabilitation, 35, 127-159.

Florida Department of Corrections (FDOC). (2003). Recidivism report: Inmates released from Florida prisons, July 1995 through June 2001. Tallahassee, FL: Florida Department of Corrections.

Florida Department of Corrections. (2004). Faith- and Character-Based Institutions. Procedure No. 506.032. May 3.

Florida Department of Corrections. (2007a). Hillsborough Correctional Institution. Retrieved May 25, 2007, from http://www.dc.state.fl.us/facilities/region3/529.html

Florida Department of Corrections. (2007b). Lawtey Correctional Institution. Retrieved May, 25, 2007, from http://www.dc.state.fl.us/facilities/region2/255.html

Florida Department of Corrections. (2008a). Chaplaincy services quick facts. Retrieved March 15, 2008, from http://www.dc.state.fl.us/oth/faith/Facts.html

Florida Department of Corrections. (2008b). Faith- and Character-Based Initiative. January 2008 update. Retrieved March 15, 2008, from http://www.dc.state.fl.us/oth/faith/stats.html

Florida Department of Corrections. (2008c). Faith- and Character-Based Institutions (FCBIs). Retrieved March 15, 2008, from http://www.dc.state.fl.us/oth/faith/ci.html

Henriques, D. B., & Lehren. A. (2006, December 10). Religion for a captive audience, paid for by taxes. The New York Times.

Hercik, J., Lewis, R., Myles, B., Gouvis, C., Zweig. J., Whitby, et al. (2005). Development of a guide to resources on faith-based organizations in criminal justice (NCJRS 209350). Washington, DC: U.S. Department of Justice.

Hercik, J. (2004). Rediscovering compassion: An evaluation of Kairos Horizon Communities in Prison. Fairfax, VA: Caliber Associates.

Johnson, B.R., & Larson, D. B. (2003). The InnerChange Freedom Initiative: A preliminary evaluation of a faith-based prison program. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

260

Johnson, B. R., Tompkins, R. B., & Webb, D. (2002). Objective Hope. Assessing the effectiveness of faith-based organizations: A review of the literature. Philadelphia: University of Pennsylvania, Center for Research on Religion and Urban Civil Society.

Kerley, K. R., Matthews, T. L., & Blanchard, T. C. (2005). Religiosity, religious participation, and negative prison behaviors. Journal for the Scientific Study of Religion, 44, 443–457.

La Vigne, N., Brazzell, D., & Small, K. (2007). Evaluation of Florida’s faith- and character-based institutions. Washington, DC: The Urban Institute.

McDaniel, C., Davis, D.H., & Neff, S. A. (2005). Charitable Choice and prison ministries: Constitutional and institutional challenges to rehabilitating the American penal system. Criminal Justice Policy Review, 16, 164–189.

Mears, D. P., Roman, C. G., Wolff, A., & Buck, J. (2006). Faith-based efforts to improve prisoner reentry: Assessing the logic and evidence. Journal of Criminal Justice, 34, 351–367.

National Institute of Corrections Information Center (NICIC). (2005). Residential faith-based programs in state corrections. Washington, DC: U.S. Department of Justice.

O’Connor, T. P. (2004). What works: Religion as a correctional intervention: Part I. Journal of Community Corrections, 14, 11–27.

O’Connor, T.P. (2005). What works: Religion as a correctional intervention: Part II. Journal of Community Corrections, 14, 4–6, 20–26.

O’Connor, T. P, Duncan, J., & Quillard, F. (2006). Criminology and religion: The shape of an authentic dialogue. Criminology and Public Policy, 5, 559–570.

Sumter, M. (2000). Religiousness and post-release community adjustment (NCJRS 184508).Washington, DC: U.S. Department of Justice.

Sumter, M. (2006). Faith-based prison programs. Criminology and Public Policy, 5, 523–528.

White House. (2008). The quiet revolution: The President’s Faith-Based and Community Initiative: A seven-year progress report. Washington, DC: The White House Office of Faith-Based and Community Initiatives.

261

CONSTITUTIONAL CHANGE AND RESPONSIBILITIES OF GOVERNANCE PERTAINING TO THE FAITH-BASED

AND COMMUNITY INITIATIVE

Ira C. Lupu Robert W. Tuttle

Thirty years ago, a governmental initiative designed to encourage the participation of houses of worship in government-financed social services would have been unthinkable. At that time, many faith-based social welfare providers received public funding, but those providers were generally large, highly secularized entities (Monsma, 1996; Saperstein, 2003). For years, prominent religious denominations had chosen to deliver social services through separately incorporated and specialized entities, such as Catholic Charities and Lutheran Social Services. This service delivery structure permitted social service professionals to carry out their functions without day-to-day oversight by religious leaders. The structure also insulated houses of worship from the responsibilities and potential liabilities associated with the delivery of social services.

As a constitutional matter, the First Amendment’s Establishment Clause, which provides that “Congress shall make no law respecting the establishment of religion,” stood as an impediment to government financing of houses of worship and other institutions that engaged in religious instruction. The Supreme Court had interpreted the Clause to prohibit government funds from flowing directly to “pervasively sectarian” entities (Hunt v. McNair, 1973), a category that encompassed houses of worship, religious elementary and secondary schools, and colleges with a strong religious identity.

Between 1988 and 2000, however, the relevant constitutional law shifted, and policies on public aid to faith-based service providers began to follow suit. In Bowen v. Kendrick (1988), the Supreme Court upheld legislation that explicitly allowed religious organizations to receive funding as service providers for a program aimed at adolescent sexuality. Following the Bowen decision, courts started to relax the constitutional restrictions governing expenditures to religious entities. Themes of equal access to public resources by religious actors began to emerge in the legal and political culture, as exemplified by the Charitable Choice provisions of the welfare reform legislation, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. These provisions include (1) a mandate on participating states to treat religious entities on the same basis as secular service providers, (2) a promise that participating religious organizations can retain their religious character and identity, (3) a requirement that all providers respect the religious liberty of beneficiaries, (4) a prohibition on direct government funding of

263

“sectarian worship, instruction, or proselytization,” and (5) an affirmation that participating religious organizations retain their right to make religion-based employment decisions.

When George W. Bush began his presidency, the legal ban on grants to pervasively sectarian entities had been effectively eliminated (Mitchell v. Helms, 2000). The Supreme Court’s decision in the Cleveland school voucher case (Zelman v. Simmons-Harris, 2002), which upheld against constitutional challenge a program of voucher-based financial support for education in private schools (some with a strong religious identity), paved the way for greater latitude in the involvement of faith-based organizations in government-financed programs. Nevertheless, important constitutional restrictions on government support for religious activity remain in effect.

Immediately on taking office in January 2001, President Bush issued executive orders that set in motion the Faith-Based and Community Initiative (FBCI) (Exec. Orders 13,198–13,199). These orders created the White House Office of Faith-Based and Community Initiatives (WHOFBCI) and established centers for this initiative in the most prominent grant-making agencies in the federal government. The President ordered the agency centers to find and eliminate unnecessary obstacles to the participation of faith-based and community organizations in government-financed social services. In 2003 and thereafter, these agencies promulgated a series of regulations designed to eliminate such impediments and ensure the equal treatment of religious organizations seeking to partner with federal or state agencies in the delivery of social services. For example, the U.S. Department of Health and Human Services (HHS) issued a regulation focused on the equal treatment of religious organizations that participate in the agency’s funding programs, and many federal agencies have similar regulations (for additional examples, see the Table of Legal Materials in Appendix A).

From the beginning of this effort, however, the FBCI has faced the formidable task of translating a set of emerging constitutional principles into workable regulatory concepts. This initiative rests on a premise of neutrality or nondiscrimination between secular and religious providers; in other words, faith-based providers are equal to secular service providers in eligibility to participate in the delivery of government-financed services. But the relevant constitutional law still embodies remnants of a premise in tension with complete neutrality; namely, religious activities are constitutionally different from secular activities, even if both are undertaken for the same public purposes. The Establishment Clause of the First Amendment continues to impose constraints on public financing of religious activity, and many state constitutions have similar limitations.

264

The existing literature on the legal parameters surrounding the FBCI has tended to take one of two forms. Much of the literature emphasizes the social, religious, cultural, political, and administrative milieus in which this initiative developed, and does not delve deeply into legal considerations (Monsma, 1996; Farris, Nathan, & Wright, 2004; Ryden & Polet, 2005; DiIulio, 2007). More traditional legal literature either defends the FBCI’s basic premise of “the level playing field” (Esbeck, 1997) or is critical of the initiative for being insensitive to traditional concerns of church-state separation (Saperstein, 2003; Gilman, 2002, 2007a).

However, a more comprehensive approach to legal analysis of the FBCI focuses on the rate and direction of legal change in the relevant constitutional norms, the details of the implementation to the Initiative and the constitutional constraints that surround the FBCI, and the litigation associated with this Initiative (Lupu & Tuttle, 2002, 2003, 2004, 2005a, 2005b, 2006, 2007). This paper addresses the regulatory response of various government agencies to the relationship between the FBCI’s agenda and the changing legal environment. After briefly mapping the key developmental points in the relevant constitutional law, the discussion focuses on the particulars of regulatory language as put into place by the agencies of the U.S. government and considers the litigation efforts that have occurred in conjunction with this initiative. Some of the litigation may be traced, at least in part, to inadequacies in the FBCI’s regulatory reform. In addition, the paper provides recommendations for federal and state regulatory language that could be used to facilitate the FBCI while keeping it within constitutional bounds. The paper does not systematically evaluate the less formal documents and processes on which the federal government frequently relies to guide grant recipients.

THE EVOLVING CONSTITUTIONAL ENVIRONMENT

The regulatory reforms and the litigation challenges prompted by the FBCI can best be understood in light of the evolving principles of First Amendment law. For over six decades, the Supreme Court has consistently interpreted the Establishment Clause of the First Amendment to impose some constraints on public financing of religious entities (Everson v. Board of Ewing Township, 1947). Although the interpretive details have changed, the Court’s guiding principle throughout has been that the government may not promote or directly subsidize religious worship or instruction in a particular faith. When the government engages in or directly supports religious indoctrination, it violates the core understanding of the First Amendment ban on “establishment of religion.” This prohibition, called the “no-indoctrination” principle, is absolute. Unlike other values protected by the First Amendment, the ban on religious establishment is never balanced against the social benefits that would arise from a violation of the principle. Many state constitutions have similar limitations (Lupu & Tuttle, 2002).

265

From the early 1970s until the late 1990s, the Supreme Court implemented that principle with a sharply defined rule that government could not make direct grants to “pervasively sectarian” entities, such as houses of worship, religious schools, and even social welfare organizations that had an overtly religious character. The Court reasoned that grants to such organizations would inevitably support the mission of religious indoctrination because the government was incapable of ensuring that the organization would use public funds solely for secular activities.

Between 1981 and 2002, however, three new and highly relevant constitutional principles emerged. The first principle involves the Court’s gradual abandonment of the overly broad rule that pervasively sectarian entities may not participate in social welfare programs and replacement of that rule with a narrower no-indoctrination rule that the government may not directly finance religious activities (Agostini v. Felton, 1997; Mitchell v. Helms, 2000). At the heart of this shift is the Court’s determination that faith-based entities will not inevitably divert public resources to serve their religious mission, but instead may be trusted to comply with restrictions on how public aid may be used. Such trust is conditional; the Court ruled in Mitchell that aid to religious entities must incorporate adequate safeguards, including monitoring, against the diversion of public funds to religious use.

The second principle concerns “beneficiary choice.” When the funding program is structured to permit beneficiaries a genuine and independent choice among secular and religious providers, and public funds are directed to the provider only as a result of such choices by beneficiaries, religious organizations may use these funds for religious indoctrination as an element of the relevant service (Zelman v. Simmons-Harris, 2002). For example, school voucher programs may permit participating schools to include prayer or religious instruction as an element of the education of voucher students, so long as students and their families have a genuine choice among religious and secular schools and schools only receive funds as a result of the exercise of such a choice. This financing structure is typically labeled indirect funding because of the role of the beneficiary as an intermediary in distribution of public funds to religious entities. Programs that fall under the principle of beneficiary choice thus are outside the scope of the no-indoctrination principle because the beneficiary, rather than the government, is responsible for any religious indoctrination that occurs within the funded program.

The third principle involves contexts in which the government is providing a forum for speech rather than grants for the performance of services; in such situations the Court began to recognize claims of nondiscrimination, or equal access, by religious persons or causes (Widmar v. Vincent, 1981; Lamb’s Chapel v. Center Moriches Union Free School Dist., 1993; Rosenberger v. Univ. of Virginia, 1995; Good News Club v. Milford Central School, 2001).

266

Supporters of the FBCI embrace this principle of equal access and argue that it should be extended to public grants and other forms of financial support. In Locke v. Davey (2004), however, the Court ruled that, even in a beneficiary choice program, states have broad discretion in choosing whether or not to provide financial support for religious activity. Thus, in programs involving public expenditures for services, a policy of equal access is constitutionally permissible, but not constitutionally mandatory.

Unless the three principles are understood in a textured and interactive way, there are deep and ineradicable tensions among them. An unmodified principle of equal access seems at first glance to fully support the premise, central to the FBCI, that government grant programs should provide a level playing field for competition among religious and secular service providers. Moreover, an unadorned principle of beneficiary choice appears to suggest that the government may support social service with religious content to beneficiaries who voluntarily select that kind of service, so long as the government provides an adequate range of religious and secular options to beneficiaries.

The no-indoctrination principle, however, constrains the scope of the equal access principle by prohibiting the government from directly financing programs that include explicitly religious content. Moreover, the no-indoctrination principle is not limited to cases of involuntary or coercive religious experience. Because the government may not directly pay to indoctrinate even those who freely seek such an experience, beneficiary choice programs must involve beneficiary control over the direction of funds as well as genuine choice among secular and religious service providers.

Constitutionally appropriate administration of the FBCI thus depends in significant ways on (1) a nuanced appreciation of the three principles, (2) identification of areas of reconciliation among them, and (3) careful guidance to both government grantors and potential grantees about these cross-cutting constitutional parameters.

REGULATORY REFORM AND THE IMPERFECT MESH WITH CONSTITUTIONAL CHANGE

Among the very first official acts of President Bush were his two executive orders, issued January 29, 2001, designed to create an active administrative presence for the FBCI. These orders created WHOFBCI (E.O. 13,199) and the FBCI centers within each of the six major grant-making agencies of the federal government (E.O. 13,198). Both orders emphasized the principle of equal access by affirming the concepts of a level playing field and evenhandedness

267

among religious and secular organizations. Neither order explicitly referenced the constraint imposed by the no-indoctrination principle or the opportunities created by the beneficiary choice principle. One of the chief purposes of both orders was the elimination of “unnecessary legislative, regulatory, and other bureaucratic barriers that impede effective faith-based and other community efforts to solve social problems.”

In 2001–2002, the main sticking point in the congressional deliberations on legislation related to the FBCI was the issue of religious selectivity in hiring by publicly supported religious entities (Farris, Nathan, & Wright, 2004). Many supporters of the FBCI wanted to reaffirm that such organizations would retain their current exemption, codified in Title VII of the 1964 Civil Rights Act, from the prohibition on religious discrimination in employment. Others, including some who supported most of the FBCI principles, believed that faith-based entities should not be able to limit publicly financed jobs to members of their own faith. Still others feared that the FBCI would lead to violations of the no-indoctrination principle or that the mission of religious entities would be compromised if they accepted public funds.

On December 12, 2002, the President acted unilaterally with an executive order entitled “Equal Protection of the Laws for Faith-Based and Community Organizations.” This order made the key principles of the FBCI applicable to all federal programs of social welfare funding.

In 2003–2004, through the coordination of WHOFBCI and the FBCI agency centers, all of the major granting agencies promulgated a set of regulations designed to guide the participation of faith-based organizations in the grant-making process. These regulatory reforms (see the Table of Legal Materials in Appendix A) have achieved three substantial successes. First, the regulations have clarified that houses of worship and other entities with a strong religious identity are eligible to receive public grants as social welfare service providers. Prior to the Supreme Court’s decision in Mitchell v. Helms (2000), which effectively eliminated the exclusion of pervasively sectarian entities from direct government funding, the eligibility of such organizations had been in doubt. These sets of regulations specify that participating organizations have a right to retain their religious character, including the display of religious icons, the selection of a governing board on a religious basis, and the inclusion of religious references in their mission statements.

Second, the regulations provide that service beneficiaries have the right to be free from religious coercion or discrimination in federally financed activity. No prior U.S. statute or regulation prohibited federal grantees from engaging in discrimination against beneficiaries based on religious identity or on the beneficiaries’ willingness to participate in religious activity. This is a

268

very important development, although respect for these rights could be enhanced through an affirmative obligation of faith-based grantees to inform beneficiaries of the protections. HHS regulations for substance abuse treatment services provide an example of such a notice.

Third, the regulations address the responsibility of all grantees with respect to the use of government funds to support religious activities. Nearly all federal agencies involved in the FBCI now have in place a regulation prohibiting the direct financing of “inherently religious activities, such as religious worship, instruction, or proselytization.” Grantees may engage in such activities only with private funds, and must separate those activities in time or place from activities directly funded by the government. These content restrictions have been reinforced by a variety of guidance documents and training materials developed by the WHOFBCI and federal agencies.

The emphasis on the concept of “inherently religious activities,” however, may have produced an ambiguity in the regulatory environment. The concept of inherently religious activities is derived from Bowen v. Kendrick (1988) where the Court ruled that services related to adolescent sexuality were not “inherently religious” because such services could be performed with exclusively secular content. Thus, the FBCI’s ban on direct funding for inherently religious activities is accurate to a point because the no-indoctrination principle bars public support for worship, religious instruction, and proselytizing. But the regulations do not confront the question of whether social services that have explicitly religious content are precluded by the regulation. If understood too narrowly, the regulatory proscription on direct government financing of religious instruction significantly understates the no-indoctrination principle, which bars direct government financing of any activity with “specifically religious content” (Bowen v. Kendrick, 1988).

We believe the existing regulatory formula creates ambiguity with respect to direct funding of social services that have religious content. Some social services may be delivered using religious content, even if such content is not a necessary part of the service. For example, a grantee may teach the skills and behaviors necessary for success in the workplace, the value of marriage, the benefits of sexual abstinence outside of marriage, or the benefits of a drug-free lifestyle. The provision of these services may be in entirely secular terms or in terms that include explicitly religious language and beliefs. While it is possible that federal officials might provide sufficient warning and guidance, the regulations themselves do not warn grantors and grantees that the inclusion of specifically religious content in the delivery of such services may put the granting agency in conflict with the no-indoctrination principle and expose both the public grantor and the private grantee to legal consequences.

269

At times, federal agencies have provided appropriate explanation that the prohibition on direct funding of “religious instruction” includes social service materials that contain religious content. For example, earlier this year, the Administration for Children and Families (ACF) issued a memorandum to grantees under the Healthy Marriage Initiative. The memorandum explained that curricular materials must be “neutral with respect to religion,” and provided examples of curricula that would not qualify for direct government support because the materials included religious content. Another example of appropriate guidance can be found in the recent memorandum—issued by ACF in the wake of the settlement in ACLU of Massachusetts v. Leavitt—on the proper and improper ways to spend grant funds for programs that promote sexual abstinence by minors.

An additional, though related, ambiguity may arise from the regulatory guarantee that faith-based organizations may receive grants and still retain their religious character. Even though the regulations define “religious character” in terms associated with organizational identity (acceptability of religious board members, maintenance of religious content in mission statement, etc), it is possible that organizations may misinterpret the guarantee as a representation that religious content in their government-financed services is acceptable if such content reflects the entity’s religious character.

The ambiguity of the regulatory focus on “inherently religious activities” is especially problematic when applied to the direct funding of social services that aim at personal transformation. If faith-based social service agencies deliver food, shelter, or other material goods, it is usually not difficult for them to segregate those goods from religious activity and to ensure that government funds do not pay for the latter. In such a context, government funds can be used exclusively to support secular goods and services, and religious activity can be offered, as the federal regulations uniformly require, at a “separate time or place” not subsidized by government funds. In contrast, transformative services, such as those aimed at substance abuse, sexual abstinence among unmarried minors, and rehabilitation of those incarcerated for crime, may be much harder to segregate into secular and religious components. Most of the successful litigation (discussed below) against government-financed faith-based services has arisen in the context of transformative services.

Uncertainty about the constitutionally acceptable use of direct grants also presents a difficult problem in the context of grants designed to build the capacity of faith-based organizations to provide social services. Such grants have been a central building block of the FBCI through the Compassion Capital Fund (CCF). Depending on the activities funded by the capacity-building grants, it can be difficult to segregate secular and religious activities. If a recipient of such grants

270

provides both religious and secular services, aid for building the organization’s capacity—through computer technology, for example—is likely to benefit the organization’s religious programs as well as its secular programs, thus making the aid vulnerable to constitutional challenge. The one decided case on the subject upheld a capacity-building grant to a marriage-strengthening program, but did so only in light of a specific finding that the grantee had removed the religious content from all the activities that benefited from federal support for the grantee’s capacity. (Christianson v. Leavitt, 2007). Thus, the public funding for capacity-building supported exclusively secular services.

Moreover, CCF grants often involve private intermediaries empowered by the government to make subgrants. The Supreme Court has ruled that, like the government, these subgrantors are prohibited from directly subsidizing religious indoctrination (Bowen v. Kendrick, 1988). In the early years of the FBCI, the federal agencies may not have been sufficiently sensitive to the problems of necessary guidance for either intermediaries or subgrantees. The concern remains that the federal regulations have generally repeated the formula by which direct grants may not be used for inherently religious activities.

An additional concern related to the current regulations is the existing legal requirement that the government monitor the content of government-financed social services provided by faith-based entities. The Supreme Court has ruled that such monitoring, which should not and need not be intrusive, is necessary to ensure compliance with the no-indoctrination principle; that is, to protect against unlawful diversion of public money to religious activity (Mitchell v. Helms, 2000; Agostini v. Felton, 1997). While the grantees are subject to the same federal or state monitoring imposed on any grantee, the current FBCI regulations prohibit the imposition of any requirement that faith-based organizations be treated differently than secular organizations for purposes of regulatory compliance. Although the emphasis on equal treatment is salutary, the regulations may not adequately highlight the government’s continuing constitutional obligation to provide particularized monitoring that focuses on the risk of diversion of public funds to religious activities. In a 2006 study of programs that provide funding for faith-based and community organizations, the Government Accountability Office found that federal agencies gave auditors little guidance in assessing whether grantees complied with restrictions on use of direct funds for religious activities (GAO, 2006).

Even with respect to indirect funding, where constitutional restrictions are looser and room for inclusion of religious content correspondingly greater, federal programs could provide better regulatory language for government officials and faith-based grantees. Apart from regulations implementing statutory Charitable Choice provisions, regulations promulgated by the federal

271

agencies do not define indirect funding. Instead, the agency-wide FBCI regulations proscribe certain religious content in situations of direct funding and indicate that such a proscription does not apply to a program of indirect funding.

Although this omission has not appeared to cause the same degree of uncertainty as the proscription on direct financing of inherently religious activities, potential grantors and grantees could benefit from specific regulations with respect to indirect funding. (Some agencies have developed detailed resources that provide clarity about the constitutional parameters of indirect funding. See for example the Department of Labor (2005) resource “Partnering with Integrity.”)

The phrase “indirect funding” is drawn from the Cleveland school voucher decision (Zelman v. Simmons-Harris, 2002) and other Supreme Court cases involving policies that empower individual beneficiaries to select a religious entity as the provider of publicly supported services (Zobrest v. Catalina Foothills Sch. Dist., 1993; Witters v. Wash. Dept. of Services for the Blind, 1986). The Court’s decisions upholding such programs emphasized the ability of beneficiaries to make a “genuine and independent choice” among secular and religious alternatives. Thus far, the agency-wide FBCI regulations for implementing programs of beneficiary choice have only indicated that the programs must offer beneficiaries a choice to which the beneficiaries do not have religious objections. Those regulations have not taken the constitutionally mandated next step of requiring that voucher programs offer a religiously neutral menu of service providers, including a secular alternative.

Moreover, federal regulations have not yet made clear the requirement that indirect aid must in some fashion pass through the control of the beneficiary, though not necessarily by the mechanism of vouchers or other tangible instruments (Freedom from Religion Foundation, Inc. v. McCallum, 2003). In other words, the concept of beneficiary choice means more than the simple fact of voluntary selection by beneficiaries among a number of providers. The regulations should state that the beneficiary must be empowered to designate the provider as a recipient of government funds and to stop the flow of those funds by withdrawing from that provider.

In addition, the regulations might fruitfully sharpen the concept of beneficiary choice by greater attention to particular service contexts. What may constitute an adequate mechanism for choice when parents choose elementary or secondary schools (Zelman v. Simmons-Harris, 2002) may be insufficient when the relevant service involves people suffering from drug addiction, with accompanying impairment of cognitive capacity, or institutionalized foster care of minor children arranged by state officials (Teen Ranch, Inc. v. Udow, 2007). In cases of the latter type,

272

involving state officials rather than parents acting on behalf of minors, it may be that no one is competent to exercise beneficiary choice in favor of religious programming.

Finally, despite substantial efforts to involve the states in the efforts associated with the FBCI, the regulatory analysis and communication of various legal doctrines of federalism could be improved for this initiative. For example, federal agencies have at times confusingly explained the relationship between federal law and state law, especially with respect to the highly controversial issue of religious selectivity in hiring by faith-based organizations. An exemption from federal restrictions on faith-based hiring does not automatically create an exemption for federal grantees from state law restrictions on faith-based hiring decisions.

Moreover, the states themselves—even those with their own active faith-based and community initiatives—generally have not taken on the necessary task of supplying state-level regulatory content to granting agencies and potential grantees. To be successful at the state and local levels, the FBCI needs to include a legal and bureaucratic program designed to institutionalize its premises. Whether the cause is legal uncertainty or bureaucratic inertia, the states, which are the locus of most grant-making activity, have made virtually no progress in developing their own regulatory frameworks, even with respect to specific questions raised by state constitutional law. In some states, these state constitutional questions are substantial and include limits on the use of even indirect public aid for religious activity (Lupu & Tuttle, 2002).

THE LITIGATION SURGE PRODUCED BY THE FBCI

Concerns about regulatory precision and transparency are not merely academic. Since the inception of the FBCI, government funding of faith-based social services has generated a significant volume of litigation and there has been a considerable increase in the number of lawsuits and judicial decisions about publicly financed, faith-based social services in the past 6 years. Most of these lawsuits have been initiated by public interest groups that are committed to strong principles of church-state separation (see the Table of FBCI-Related Litigation in Appendix B). Freedom from Religion Foundation, Inc. (FFRF) has been the most active instigator of this sort of litigation. Americans United for Separation of Church and State, the American Civil Liberties Union, the American Jewish Congress, and other groups have played a similar role.

The Table of FBCI-Related Litigation lists all of the cases brought so far, categorized by outcome: Victories for Challengers to Government Action (7), Victories for Defenders of

273

Government Action (9), Settlements and other Terminations Before Decision (7), and Lawsuits Still Pending (2). These cases involve a wide variety of social services, including treatment for substance abuse, sexual abstinence for unmarried minors, work training, prisoner rehabilitation, mentoring the children of prisoners, education in pastoral care for nurses, custodial foster care for troubled teenagers, chaplaincies for public employees and patients in public hospitals, and the provision of shelter for the homeless. In a few of these cases early in the FBCI, the evidence suggested that the grantee had received inadequate guidance, which led to settlements in favor of the challengers or judicial rulings against the government (see, for example, FFRF v. Montana Office of Rural Health, 2004; ACLU of Massachusetts v. Leavitt, 2006).

An elaborate body of law has now developed on the subject of partnerships between government and faith-based organizations in the delivery of social services, although few lawyers are aware of this as a defined field. Most of the victories and favorable settlements for the challengers to government action and several of the government victories as well (Teen Ranch v. Udow, 2007; Christianson v. Leavitt, 2007) have involved a strong invocation of the no-indoctrination principle. If grantees use direct government support to finance the interweaving of religious themes with social services, they invite successful lawsuits against the grantor and themselves. All lower court decisions have upheld the importance of this prohibition on government-financed indoctrination. In sharp contrast, the government has typically prevailed in suits asserting other kinds of claims, such as unsuccessful allegations of unlawful preference for religious over secular entities (FFRF v. Towey, 2005), unlawful religious discrimination in hiring (Lown v. Salvation Army, 2005), or improper structures of indirect financing (FFRF v. McCallum, 2003; American Jewish Congress v. Corporation for National and Community Service, 2005).

In June 2007, the Supreme Court decided an important case that challenged the general policies of the FBCI rather than specific grants made under it. In Hein v. Freedom from Religion Foundation, Inc., the Court ruled that federal taxpayers lacked standing to complain about the constitutionality of discretionary expenditures made by federal executive agencies to promote the FBCI. The Hein ruling rests on a distinction between legislative decisions to make expenditures explicitly in support of religious causes or entities and discretionary executive decisions to spend in service of these same ends. Taxpayers may now challenge in the federal courts only spending decisions that have been explicitly authorized by legislatures. The Hein decision may well slow the pace of challenges to particular grants under the FBCI because some of these grants are not made pursuant to explicit legislative authority; however, Hein will not eliminate such challenges (Lupu & Tuttle, 2008).

274

IMPLICATIONS

The FBCI resides in a constitutional and regulatory milieu that requires detailed regulations and careful guidance to be provided to grantors and grantees. Federal agencies have made some progress in this regard, but more needs to be done. This analysis leads to a number of recommendations that may facilitate a more administratively successful integration of the FBCI with the present contours of the relevant law:

1. The regulations should be revised to clarify the full scope of the prohibition on direct funding of activity that has explicitly religious content. Such clarification will be most useful if it specifically identifies situations, such as faith-intensive treatment programs for substance abuse, in which grantees are likely to run afoul of the no-indoctrination principle.

2. Federal agencies should expand their regulatory concern to indirect financing, with particular attention to the adequacy of secular alternatives and the precise mechanisms of beneficiary choice.

3. Federal agencies should provide greater clarity about the relationship between federal and state law, especially regarding the preemptive (or nonpreemptive) effect of exemption from federal policies of nondiscrimination in employment.

4. Federal agencies should impose a constitutionally sensitive system of monitoring to ensure that grantees do not divert funds to religious activities.

5. Federal agencies should continue to provide improved guidance for private intermediaries, especially with respect to capacity-building grants.

6. Federal agencies should continue to make the protections for beneficiaries more robust, by requiring providers to inform beneficiaries that they have the right not to participate in religious activities and that their right to receive services does not depend on such participation.

7. States and localities should continue to undertake efforts to educate grant officers and grantees, actual and prospective, concerning the relevant restrictions under both state and federal constitutional law.

If these recommendations are implemented, it is likely that fewer lawsuits will be filed challenging grants to faith-based organizations, more grantors and grantees will be willing to participate actively in the FBCI, and the rights of beneficiaries will be fully protected. Under such conditions, the attractive promise of a level playing field for religious and secular organizations, and the rich involvement of both, can be realized in ways that are consistent with constitutional norms and traditions.

275

ACKNOWLEDGMENTS

The authors thank David Wright of the Roundtable for his very helpful comments on an earlier draft of this paper.

APPENDIX A: TABLE OF LEGAL MATERIALS

Legislative and Regulatory Materials Personal Responsibility and Work Opportunity Act of 1996, Pub. L.. No. 104-193, 110 Stat. 2105 (codified as amended in various sections of 42 U.S.C., including 42 U.S.C. sec. 604a et seq. (1996)

Exec. Order 13,198, 3 C.F.R. 750 (2002)

Exec. Order 13,199, 3 C.F.R. 752 (2002)

Exec. Order 13,279, 3 C.F.R. 258 (2003)

U.S. Department of Agriculture, Equal Opportunity for Religious Organizations, 7 C.F.R. Part 16 (2004)

U.S. Department of Education, Participation in Education Department Programs by Religious Organizations; Providing for Equal Treatment of all Education Program Participants, 34 C.F. R. Parts 74, 75, 76, and 80 (2004)

U.S. Department of Health & Human Services, Participation in Department of Health & Human Services Programs by Religious Organizations; Providing for Equal Treatment of all Department of Health & Human Services Program Participants, 45 C.F.R. Parts 74, 87, 92, and 96 (2004)

U.S. Department of Housing and Urban Development, Equal Participation of Treatment for Faith-Based Organizations, 24 C.F.R. Parts 5 and 570 (2004)

U.S. Agency for International Development, Participation by Religious Organizations in U.S. AID Programs, 22 C.F.R. Parts 202, 205, 211, and 226 (2004)

U.S. Department of Justice, Equal Treatment for Faith-Based Organizations, 28 C.F.R. 38.1–38.2 (2004)

U.S. Department of Labor, Equal Treatment in Department of Labor Programs for Religious Organizations; Protection of Religious Liberty pf Department of Labor Social Service Providers and Beneficiaries, 29 C.F.R.2.2.32 (2004)

U.S. Department of Veterans Affairs, VA Homeless Providers Grant and Per Diem Program; Religious Organizations, 38 C.F.R. Part 61 (2004)

276

Selected Administrative Resources Government Accountability Office, Faith-Based and Community Initiative: Improvements in Monitoring Grantees and Measuring Performance Could Enhance Accountability (2006)

U.S. Department of Health & Human Services, Administration for Children & Families, Office of Family Assistance, “Curricular Use and Compliance with 45 CFR Part 87” (January 23, 2008)

U.S. Department of Labor, Employment and Training Administration, “Training and Employment Guidance Letter 1-05” (July 6, 2005)

White House Office of Faith-Based and Community Initiatives, Guidance to Faith-Based and Community Organizations on Partnering with the Federal Government (2006)

White House Office of Faith-Based and Community Initiatives, Promoting Equal Treatment: A Guide for State & Local Compliance with Federal Regulations (2006)

Decided Cases Bowen v. Kendrick, 487 U.S. 589 (1988)

Everson v. Bd. of Educ., 330 U.S. 1 (1947)

Flast v. Cohen, 392 U.S. 83 (1968)

Freedom from Religion Foundation, Inc. v. McCallum, 324 F. 3d 880 (7th Cir. 2003)

Good News Club v. Milford Central School, 533 U.S. 98 (2001)

Hein v. Freedom from Religion Foundation, Inc., 127 S. Ct. 2553 (2007)

Hunt v. McNair, 413 U.S. 734 (1973)

Lamb’s Chapel v. Center Moriches Union Free School Dist., 508 U.S. 384 (1993)

Lemon v. Kurtzman, 403 U.S. 602 (1971)

Locke v. Davey, 540 U.S. 712 (2004)

Mitchell v. Helms, 530 U.S. 793 (2000)

Rosenberger v. Rector & Visitors of the Univ. of Virginia, 515 U.S. 819 (1995)

Widmar v. Vincent, 454 U.S. 263 (1981)

Witters v. Wash. Dept. of Services for the Blind, 474 U.S. 481 (1986)

Zelman v. Simmons-Harris, 536 U.S. 639 (2002)

Zobrest v. Catalina Foothills Sch. Dist., 509 U.S. 1 (1993)

277

APPENDIX B: TABLE OF FBCI-RELATED LITIGATION

Victories for Challengers to Government Action Americans United for Separation of Church & State v. Prison Fellowship Ministries, 509 F.3d 406, U.S. Court of Appeals for the 8th Circuit, 2007 (affirming the invalidation of faith-based prison program, though reversing an order that Prison Fellowship Ministries repay the State of Iowa for unconstitutional programming)

American Civil Liberties Union Foundation v. Blanco, 2007 U.S. Dist. Lexis 74590, U.S. District Court for the Eastern District of Louisiana, 2007 (invalidating earmarked grants, without specified purposes or safeguards against diversion to religious activities, to named houses of worship)

Community House, Inc. v. City of Boise, 463 F.3d 1118, U.S. Court of Appeals for the 9th Cir. 2006 (affirming decision invalidating below-market lease from city to operator of homeless shelter because operator was engaged in religious indoctrination and sex discrimination)

Bush v. Holmes, 919 S. 2d 392, Supreme Court of Florida, 2006 (invalidating, on state constitutional grounds, a voucher scheme for sending children to private schools, religious and secular)

Freedom from Religion Foundation, Inc. v. Towey, 2005 U.S. Dist. Lexis 39444, U.S. District Court for the Western District of Wisconsin, 2005 (invalidating HHS grant to MentorKids USA because of religious content in the services provided)

Freedom from Religion Foundation, Inc. v. Montana Office of Rural Health, 2004 U.S. Dist. Lexis 29139, U.S. District Court for the District of Montana, 2004 (invalidating grants for the support of education in the practice of religion-based parish nursing)

Freedom from Religion Foundation, Inc. v. McCallum & Faith Works Milwaukee, 179 F. Supp.2d 950, U.S. District Court for the Western District of Wisconsin, 2002 (invalidating grant for the direct support of faith-based residential service for the treatment of substance-addicted welfare recipients)

Victories for Defenders of Government Action Hein v. Freedom from Religion Foundation, Inc., 127 S. Ct. 2553 (2007) (dismissing taxpayer challenge to executive expenditures promoting the FBCI)

Teen Ranch, Inc. v. Udow, 479 F.3d 403, U.S. Court of Appeals for the 6th Circuit, 2007 (affirming a decision upholding the exclusion from state assistance of faith-based residential service facility for teens with problems of substance abuse)

Christianson v. Leavitt, 482 F. Supp. 2d 1237, Secretary of Health & Human Services, U.S. District Court for the Western District of Washington, 2007 (dismissing suit against capacity-

278

building grant by HHS to faith-based marriage support group, on the grounds that the grant supported only secular activity)

Freedom from Religion Foundation, Inc. v. Nicholson, Secretary of the Department of Veterans’ Administration, 469 F. Supp. 2d 609, U.S. District Court for the Western District of Wisconsin, 2007 (dismissing suit against hospital chaplaincy program of the Department of Veterans’ Administration)

Freedom from Religion Foundation, Inc. v. Towey, 2005 U.S. Dist. Lexis 39444, U.S. District Court for the Western District of Wisconsin, 2005 (upholding, against claim of religious discrimination, subgrants from Emory University to faith-based health providers)

Lown v. Salvation Army, Inc., & Commissioner, New York City Administration for Children’s Services, 393 F. Supp. 2d 223, U.S. District Court for the Southern District of New York, 2005 (dismissing claim that religious selectivity in employment, by publicly supported Salvation Army, violated federal or state law)

American Jewish Congress v. Corporation for National and Community Service, 399 F.3d 351, U.S. Court of Appeals for the District of Columbia Cir., 2005 (rejecting constitutional challenge to system of indirect public financing for training of teachers placed at religious schools in AmeriCorps program)

Freedom from Religion Foundation, Inc. v. McCallum, 324 F.3d 880, U.S. Court of Appeals for the 7th Cir. 2003 (rejecting constitutional challenge to indirect payment for placement of drug offenders at faith-based residential treatment center for substance abuse)

American Jewish Congress v. Bost, U.S. District Court for the Southern District of Texas, 2002 (refusing to order restitution of public funds by operator of faith-based jobs program that had unlawfully used public monies for support of religious activity)

Settlements and other Terminations Before Decision Freedom from Religion Foundation, Inc. v. Mitchell Roob, U.S. District Court for the Southern District of Indiana, 2007 (withdrawal of suit after state terminated the employment of a chaplain hired to minister to employees of state social service department)

Geneva College v. Chao, Secretary of Labor, U.S. District Court for the Western District of Pennsylvania, 2007 (suit alleging unlawful interference with college’s religious autonomy in hiring settled after state and federal authorities withdrew objection to college’s use of publicly supported employment search Web site)

Freedom from Religion Foundation, Inc. v. Alberto Gonzales, U.S. District Court for the Western District of Wisconsin, 2006 (withdrawal of suit after U.S. Department of Justice suspended its request for proposals to provide faith-based programming in federal prisons)

American Civil Liberties Union of Massachusetts v. Leavitt, Secretary of Health & Human Services, U.S. District Court for the Eastern District of Massachusetts, 2006 (settlement of suit

279

against HHS for grant to “Silver Ring Thing,” a faith-based program promoting sexual abstinence among unmarried teens)

Freedom from Religion Foundation, Inc. v. Minnesota Faith/Health Consortium, Alberto Gonzales, U.S. District Court for the District of Minnesota, 2005 (withdrawal of suit after University of Minnesota altered curriculum in its program on the relationship between faith and health)

Bellmore v. Georgia Department of Human Resources & United Methodist Children’s Home, Georgia Superior Court, 2004 (settlement of employment discrimination suit in light of revision of state policy on religious discrimination in employment by state-supported faith-based organizations)

American Civil Liberties Union of Louisiana v. Foster, 2002 U.S. Dist. Lexis 13778, U.S. District Court for the Eastern District of Louisiana, 2002 (preliminary injunction, followed by settlement of suit against Governor of Louisiana with respect to program of state support for faith-based programs promoting sexual abstinence among unmarried teens)

Lawsuits Still Pending Freedom from Religion Foundation, Inc. v. Bjergaard & Dakota Boys and Girls Ranch, U.S. District Court for the District of North Dakota, 2007 (suit alleging unlawful state support for religion in faith-based group foster home)

Laskowski v. University of Notre Dame & Spellings, U.S. Secretary of Education, U.S. Court of Appeals for the 7th Circuit, appeal pending, 2008 (suit alleging unlawful subgrants of public monies by University of Notre Dame to religiously affiliated colleges in federal program of teacher training)

REFERENCES

DiIulio, J. J., Jr. (2007) A centrist blueprint for America’s faith-based future. Wildavsky Forum 2007.

Esbeck, C.H. (1997). A constitutional case for governmental cooperation with faith-based social service providers. 46 Emory L. J. 1.

Farris, A., Nathan, R., & Wright, D. (2004, December). The expanding administrative presidency: George W. Bush and the Faith-Based Initiative. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Formicola, J. R., Segers, M. C., & Weber, P. (2003). Faith-based initiatives and the Bush administration: The good, the bad, and the ugly. Lanham, MD: Rowman & Littlefield.

Gilman, M.E. (2002) Charitable Choice and the accountability challenge: Reconciling the need for regulation with the First Amendment’s religion clauses. 55 Vanderbilt L. Rev. 797.

280

Gilman, M. E. (2007) Fighting poverty with faith: Reflections on ten years of Charitable Choice. 10 Journal of Gender, Race, & Justice 395.

Gilman, M. E. (2007). If at first you don’t succeed, sign an executive order: President Bush and the expansion of Charitable Choice. 15 William & Mary Bill of Rights J. 1103.

Lupu, I. C., & Tuttle, R. W. (2002, December). Government partnerships with faith-based service providers: The state of the law. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Lupu, I. C., & Tuttle, R. W. (2003, December). Partnerships between government and faith-based providers: Annual report on the state of the law, fall, 2003. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Lupu, I. C., & Tuttle, R. W. (2004, December). The state of the law, 2004: Partnerships between government and faith-based organizations. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Lupu, I. C., & Tuttle, R. W. (2005). The faith-based initiative and the constitution. 55 DePaul L. Rev. 1.

Lupu, I. C., & Tuttle, R. W. (2005, December). The state of the law, 2005: Legal developments affecting partnerships between government and faith-based organizations. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Lupu, I. C., & Tuttle, R. W. (2006, December). The state of the law, 2006: Legal developments affecting government partnerships with faith-based organizations. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Lupu, I. C., & Tuttle, R. W. (2007, December). The state of the law, 2007: Legal developments affecting government partnerships with faith-based organizations. The Roundtable on Religion and Social Welfare Policy, Nelson A. Rockefeller Institute of Government, SUNY.

Lupu, I. C., & Tuttle, R. W. (in press). Ball on a needle: Hein v Freedom of Religion Foundation, Inc. and the future of establishment clause adjudication. B.Y.U. L. Rev.

Monsma, S. V. (1996). When sacred and secular mix: Religious nonprofit organizations and public money. Lanham, MD: Rowman & Littlefield.

Ryden, D. K., & Polet, J. (2005). Sanctioning religion? Politics, law, and faith-based public services. Boulder, CO: Lynne Rienner.

Saperstein, D. (2003). Public accountability and faith-based organizations. 116 Harv. L. Rev. 1353.

281

U.S. Department of Labor, Center for Faith-Based and Community Initiatives. (2005, July 27). Partnering with integrity—what faith-based organizations, community groups, and workforce system leaders must know about U.S. Department of Labor equal treatment and religion-related regulations. Retrieved from http://www.dol.gov/cfbci/legalguidance.htm

White House. (2008, February). The quiet revolution—the president’s faith-based and community initiative: A progress report. Washington, DC: Author.

282

ABOUT THE CONTRIBUTORS

Scott W. Allard is Associate Professor in the School of Social Service Administration at the University of Chicago. He has been a Visiting Scholar at the National Poverty Center, the Institute for Research on Poverty, and the Institute for Policy Research. His primary research interests are social welfare policy, poverty, and nonprofit organizations.

Courtney Barthle is Senior Technical Specialist in the Community, Family, and Education Studies division at ICF International in Fairfax, Virginia. Her research and technical assistance efforts are directed toward welfare policy and implementation, with specific attention to child-only cases, faith-based programs, and rural issues.

Diana Brazzell is Research Associate in the Justice Policy Center at the Urban Institute in Washington, DC. Her research focuses on prisoner reentry, faith-based programming, and the impact of incarceration on families and communities.

Mitchell Brown is Director for Research at the Institute for Community Peace in Washington, DC, and Assistant Professor in the Department of Political Science at Auburn University. Her research focuses on the empowerment efforts of marginalized communities, both in the United States and abroad.

Fred De Jong is Senior Research Consultant, Community Research Institute, Johnson Center for Philanthropy and Nonprofit Leadership, Grand Valley State University in Grand Rapids, Michigan. His research examines factors that influence service delivery effectiveness and client outcomes among vulnerable populations; in particular, faith-based service delivery issues, substance abuse treatment outcomes, services to the chronically mentally ill, and health care access for low-income persons.

Robert L. Fischer is Research Associate Professor at the Mandel School of Applied Social Sciences and Co-Director of the Center on Urban Poverty and Community Development at Case Western Reserve University. He teaches evaluation methods in social science administration and nonprofit management, and conducts various evaluation projects.

Jeanette Hercik is Vice President in the Community, Family, and Education Studies division at ICF International in Fairfax, Virginia. Her research interests include poverty studies, at-risk children and families, and community capacity building.

Claudia B. Horn is President of Performance Results, Inc., a Washington, DC-based firm that provides program evaluations and technical assistance and training in outcomes measurement, community needs assessment, team-building, and strategic planning. Her areas of expertise include evaluation, education, and staff development.

Richard C. Hula is Professor and Chair of the Department of Political Science at Michigan State University. His research and teaching interests focus broadly on environmental policy and urban politics. His current work includes the impact of faith-based organizations of public service delivery, changing state/local environmental policy, and the political consequences of state level interventions into local policy arenas.

Leyla Ismayilova is a doctoral candidate in the School of Social Work at Columbia University. Her area of concentration is advanced clinical practice. Her research interests include

283

international social work, particularly in developing culturally congruent psychosocial interventions for vulnerable children and women with various health and mental health needs.

Cynthia Jackson-Elmoore is Dean of the Honors College, Associate Professor of Political Science and Social Work, and a Faculty Associate with the Global Urban Studies Program at Michigan State University. Her current research interests include nonprofit organizations and urban politics, state legislative processes, and promoting physical activity through public policy.

Byron Johnson is Professor of Sociology and Co-Director of the Institute for Studies of Religion at Baylor University. He is an authority on the scientific study of religion. His recent research examines the influence of the “faith factor” in crime reduction among at-risk youths in urban communities. His work also assesses the degree to which faith-based prison programs may impact ex-prisoners returning to society.

Nancy G. La Vigne is Senior Research Associate in the Justice Policy Center at the Urban Institute, Washington, DC. Her research interests focus on prisoner reentry, the evaluation of criminal justice programs and technologies, the spatial analysis of crime and criminal behavior, crime prevention strategies, and crime mapping.

Ira C. Lupu is the F. Elwood & Eleanor Davis Professor of Law at George Washington University Law School. He also serves as Co-Director of Legal Research for the Roundtable on Religion and Social Welfare Policy at the State University of New York. He is a nationally recognized scholar in constitutional law, particularly in the area of religion clauses of the First Amendment.

Rebecca Makar, at the time of this writing, was Research Associate for Poverty, Welfare, and Youth at ICF International in Fairfax, Virginia. She has expertise in technical assistance, with over 5 years of experience working at national organizations providing technical assistance to local communities and cities. Her major interests include youth development, religion in public life, and the impact of faith-based and community-based organizations on civil society.

Laura Reese is Director of the Global Urban Studies Program and Professor of Political Science at Michigan State University. Her research focuses on local economic development policy and politics, comparative urban politics, and public personnel administration. She has conducted large-scale policy evaluations for the Department of Commerce as well as for state and local economic development programs.

Fred M. Ssewamala is Assistant Professor of Social Work at Columbia University and Senior Fellow at the New America Foundation, Washington, DC. His research interests include international social and economic development, comparative social welfare policy, gender and development, and designing projects and programs for poverty alleviation and community development.

Robert W. Tuttle is Professor of Law and the David R. and Sherry Kirschner Berz Research Professor of Law and Religion at the George Washington University Law School. He also serves as Co-Director of Legal Research for the Roundtable on Religion and Social Welfare Policy at the State University of New York. His research focuses on relationships between religious institutions and civil government.

284

APPENDIX

THE FBCI RESEARCH COMPENDIUM PAPER SELECTION PROCESS

One of the goals of the White House Office of Faith-Based and Community Initiatives is to highlight the wide range of research on new initiatives and changes in governance that have occurred since the inception of the Faith-Based and Community Initiative (FBCI) in 2001. For this research conference, the Office of the Assistant Secretary for Planning and Evaluation and the Center for Faith-Based and Community Initiatives in the U.S. Department of Health and Human Services contracted with RTI International to issue a national call for abstracts and to convene an expert panel to review and recommend abstracts for the development of research papers.

RTI solicited abstracts focused on research studies sponsored by the federal government and private or nonprofit organizations that examine federal and state faith-based and community-based initiatives. Dissemination outlets for the call for abstracts included universities; professional associations; think tanks; social research and evaluation organizations; international, federal, and state government agencies; and federal grantee listservs. A variety of topic areas were listed as examples of areas of interest, including criminal justice and prisoner reentry, innovations in governance (e.g., capacity building, intermediary organizations), healthy marriage and healthy relationships, youth education and development, child welfare, international health, workforce development, welfare assistance, health, and substance abuse.

Because research on the FBCI is being conducted and sponsored by a variety of institutions, expert panel members were selected to represent diverse organizational affiliations, including academic institutions, foundations, and federal and state agencies. The expert panel members included

▪ Brian Grim, the Pennsylvania State University

▪ Carole Thompson, Annie E. Casey Foundation

▪ Jocelyn Whitfield, Substance Abuse and Mental Health Services Administration

▪ Christopher Winship, Harvard University

▪ Don Winstead, Florida Department of Children and Families

The expert panel scored the abstracts based on presentation of paper topic, discussion of research objectives, discussion of data, method and analysis, the strength of the literature review, and the implications for future research, policy, and practice. Each of the five areas was worth 20 points, and reviewers were asked to assign an overall descriptor and a point score for each area. There

285

was also an optional open-ended overview section where a description of strengths and weaknesses could be provided. Each abstract received a blinded review by two or three panel members.

The overarching aim of the panel was to select abstracts for papers that would present a wide range of programs and perspectives on the FBCI. The panel convened to review ratings, discuss the top-rated abstracts, and recommend abstracts for development into full papers for presentation under one of two conference track. For Track 1, 10 abstracts were selected that are national in scope. Track 1 papers are published in a compendium of conference research. For Track 2, the panel recommended 15 abstracts by emerging research scholars on government partnerships and faith-based organizations. Track 2 papers are available on the conference Web site.

286